stuck email queue - seeking automated lasso restart

classic Classic list List threaded Threaded
46 messages Options
123
Reply | Threaded
Open this post in threaded view
|

stuck email queue - seeking automated lasso restart

mail
Hello all.

Mac OS X server 10.6.8
Lasso 8.6

I am struggling lately with the dreaded stuck email queue issue. The website that has been effected has not previously experienced this issue over the past several years. There have been no recent changes and the Website did not get hit with a spike in traffic and email load. However, in the last few weeks the email queue "Failed to connect to host" issue has surfaced. It's almost as if every email that is attempted uses the maximum retries until failure with the cannot connect to host error. The only way to get things going again is to either restart Lasso service fully, or restart the site and then re-queue all messages.

ATTEMPTED SOLUTIONS

I have tried:

o Use IP numbers in place of domain names.

o Use an alternate email server.

o Send directly to SMTP Servers set to yes (but now back to No.)

o Send an email every 30 minutes in some hope that would prevent services from going to a "sleep" state - which later did not seem to be the case.

o Create an event that calls a lasso tag every 5 minutes (does an include_url to prompt a DNS lookup). This is what I took to be what is discussed as a "DNS Tickler" that others have stated cleared up DNS deafness as well as resolved the stuck email queue issue, for some at least. I have changed that though to use a lasso startup file following a post by Ke Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still hopeless in 8.6"

CONCLUSION

What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.

MONITORING

There are three components to monitoring:

1) A lasso script that searches the site database and if errors are found then the display of the file is "restart" and if nothing found it is simply "ignore". This contains the logic and criteria for checking in the email queue databases for errors and stuck email. Under development, see the SQL at the bottom for a draft of the statement to search the database.

2) Commmand line scripts that perform the stop and start of Lasso. I do not see why not to use the pre-built scripts provided for in Lasso:

        /Applications/Lasso Professional 8/Tools/startLassoService.command
        /Applications/Lasso Professional 8/Tools/stopLassoService.command

3) Monitoring tool to load the URL in part 1 and if the result is "restart" then to use the scripts in 2 to stop then start again.

For this third function, I am looking into using MONIT following the instructions at:

"Monitoring Lasso with Monit"
www.lassosoft.com/Monitoring-Lasso-with-Monit

However, it seems that most of these instructions are outdated, for Lasso 8.5

The only change I made on the monit configuration file "monitrc" is to add a line:

include /etc/monit/*

Then, I created a new sub-directory, named "monit" (within "etc"), for monit configuration files and made a new file within that named "lasso.monitrc"

Here is where I need help. What to place in the file "lasso.monitrc". This is what I have so far:

_________ start _________________
check process lasso_0
        with pidfile /tmp/LassoService8.pid
        start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
                with timeout 120 seconds
        stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
                with timeout 45 seconds

        if changed pid then alert

        if totalcpu > 15% for 2 cycles then alert
        if totalmem > 75 MB for 2 cycles then alert

        if 3 restarts within 5 cycles then timeout

        group lasso
        if totalcpu > 25% for 2 cycles then alert
        if totalmem > 100 MB for 2 cycles then alert
        depends on lasso_0

check host myserver with address www.mydomain.com
        stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
                with timeout 45 seconds
        start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
                with timeout 120 seconds

    if failed url <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso?display=none">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?display=none
       and content == 'restart'
       then restart
_________ end _________________

QUESTIONS / HELP

Note that it seems that Lasso 8.6 is already created a .pid file for Lasso service, I did not use any additional tag as has been suggested. I only have one site and, for now at least, I am planning just to restart the Lasso Service. Perhaps I need though to have a pid file created for the one default site (only one default site on this web server)?

Where can I get more details on how to create the monit configuration script?

Maybe I should give up on the Monit monitoring and expand my lasso script to do the stop and start using the third-party os_process tag? Then schedule execution of that in Lasso Admin as an event??

Michael
__________________________________________________________________

SELECT sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
FROM smtp_queue AS sq \
LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND smtp_error.error LIKE "%Failed to connect to host%"
ORDER BY sq.create_date DESC

Michael Collins, Web Systems Engineer

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

stevepiercy
Regarding:
http://www.lassosoft.com/Monitoring-Lasso-with-Monit

See:
http://lasso.2283332.n4.nabble.com/Monit-with-Lasso-8-6-on-CentOS-tp4635649p4635654.html

Note the differences between Mac and CentOS, and 8.5 and 8.6
regarding the PIDs.  Marc Pinnell made comments under
[RD_lassoService_pid] that are relevant to your observation:
http://www.lassosoft.com/tagswap/detail/RD_lassoService_pid

If you use the [RD_lassoService_pid] tag as recommended, then
you merely restart the Lasso instance, not server, and reduce
restart time.

--steve


On 2/28/13 at 10:22 AM, [hidden email] (Michael) pronounced:

>Hello all.
>Mac OS X server 10.6.8
>Lasso 8.6
>
>I am struggling lately with the dreaded stuck email queue
>issue. The website that has been effected has not previously
>experienced this issue over the past several years. There have
>been no recent changes and the Website did not get hit with a
>spike in traffic and email load. However, in the last few weeks
>the email queue "Failed to connect to host" issue has surfaced.
>It's almost as if every email that is attempted uses the
>maximum retries until failure with the cannot connect to host
>error. The only way to get things going again is to either
>restart Lasso service fully, or restart the site and then
>re-queue all messages.
>ATTEMPTED SOLUTIONS
>
>I have tried:
>
>o Use IP numbers in place of domain names.
>
>o Use an alternate email server.
>
>o Send directly to SMTP Servers set to yes (but now back to No.)
>o Send an email every 30 minutes in some hope that would
>prevent services from going to a "sleep" state - which later
>did not seem to be the case.
>
>o Create an event that calls a lasso tag every 5 minutes (does
>an include_url to prompt a DNS lookup). This is what I took to
>be what is discussed as a "DNS Tickler" that others have stated
>cleared up DNS deafness as well as resolved the stuck email
>queue issue, for some at least. I have changed that though to
>use a lasso startup file following a post by Ke Carlton on Dec
>06, 2011; 8:11pm in the thread "Email system still hopeless in 8.6"
>
>CONCLUSION
>
>What I have decided now is that the current email queue system
>CANNOT BE FIXED and but that I must live with monitoring for a
>stuck email queue and restarting Lasso automatically when that
>problem is found.
>MONITORING
>
>There are three components to monitoring:
>
>1) A lasso script that searches the site database and if errors
>are found then the display of the file is "restart" and if
>nothing found it is simply "ignore". This contains the logic
>and criteria for checking in the email queue databases for
>errors and stuck email. Under development, see the SQL at the
>bottom for a draft of the statement to search the database.
>
>2) Commmand line scripts that perform the stop and start of
>Lasso. I do not see why not to use the pre-built scripts
>provided for in Lasso:
>/Applications/Lasso Professional 8/Tools/startLassoService.command
>/Applications/Lasso Professional 8/Tools/stopLassoService.command
>
>3) Monitoring tool to load the URL in part 1 and if the result
>is "restart" then to use the scripts in 2 to stop then start again.
>For this third function, I am looking into using MONIT
>following the instructions at:
>
>"Monitoring Lasso with Monit" www.lassosoft.com/Monitoring-Lasso-with-Monit
>
>However, it seems that most of these instructions are outdated, for Lasso 8.5
>
>The only change I made on the monit configuration file "monitrc" is to add a line:
>
>include /etc/monit/*
>
>Then, I created a new sub-directory, named "monit" (within
>"etc"), for monit configuration files and made a new file
>within that named "lasso.monitrc"
>
>Here is where I need help. What to place in the file
>"lasso.monitrc". This is what I have so far:
>
>_________ start _________________
>check process lasso_0
>with pidfile /tmp/LassoService8.pid
>start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>with timeout 120 seconds
>stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>with timeout 45 seconds
>
>if changed pid then alert
>
>if totalcpu > 15% for 2 cycles then alert
>if totalmem > 75 MB for 2 cycles then alert
>
>if 3 restarts within 5 cycles then timeout
>
>group lasso
>if totalcpu > 25% for 2 cycles then alert
>if totalmem > 100 MB for 2 cycles then alert
>depends on lasso_0
>
>check host myserver with address www.mydomain.com
>stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>with timeout 45 seconds
>start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>with timeout 120 seconds
>
>if failed url <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?
>display=none
>and content == 'restart'
>then restart
>_________ end _________________
>
>QUESTIONS / HELP
>
>Note that it seems that Lasso 8.6 is already created a .pid
>file for Lasso service, I did not use any additional tag as has
>been suggested. I only have one site and, for now at least, I
>am planning just to restart the Lasso Service. Perhaps I need
>though to have a pid file created for the one default site
>(only one default site on this web server)?
>
>Where can I get more details on how to create the monit configuration script?
>
>Maybe I should give up on the Monit monitoring and expand my
>lasso script to do the stop and start using the third-party
>os_process tag? Then schedule execution of that in Lasso Admin
>as an event??
>
>Michael
>__________________________________________________________________
>
>SELECT sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.
>smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.
>smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.
>queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>FROM smtp_queue AS sq \
>LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE
>"%sending%" OR sq.queue_status LIKE "%queued%") AND
>smtp_error.id IS NOT NULL AND smtp_error.error LIKE "%Failed to
>connect to host%"
>ORDER BY sq.create_date DESC
>
>Michael Collins, Web Systems Engineer
>
>#############################################################
>This message is sent to you because you are subscribed to
>the mailing list Lasso
>[hidden email]
>To unsubscribe, E-mail to: <[hidden email]>
>Send administrative queries to  <[hidden email]>

-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
-- --
Steve Piercy               Web Site Builder              
Soquel, CA
<[hidden email]>                  <http://www.StevePiercy.com/>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Patrick Larkin-2
In reply to this post by mail
I've seemed to have mostly resolved this issue by simply scheduling a lasso file that sends me an email every three hours.  I've only had the failure message you describe once in the past several months.  When I stop getting the alert email, I check the queue.  Not sure why this works but it seems to do the job for me.

Patrick

On Feb 27, 2013, at 9:22 PM, Michael <[hidden email]> wrote:

> Hello all.
>
> Mac OS X server 10.6.8
> Lasso 8.6
>
> I am struggling lately with the dreaded stuck email queue issue. The website that has been effected has not previously experienced this issue over the past several years. There have been no recent changes and the Website did not get hit with a spike in traffic and email load. However, in the last few weeks the email queue "Failed to connect to host" issue has surfaced. It's almost as if every email that is attempted uses the maximum retries until failure with the cannot connect to host error. The only way to get things going again is to either restart Lasso service fully, or restart the site and then re-queue all messages.
>
> ATTEMPTED SOLUTIONS
>
> I have tried:
>
> o Use IP numbers in place of domain names.
>
> o Use an alternate email server.
>
> o Send directly to SMTP Servers set to yes (but now back to No.)
>
> o Send an email every 30 minutes in some hope that would prevent services from going to a "sleep" state - which later did not seem to be the case.
>
> o Create an event that calls a lasso tag every 5 minutes (does an include_url to prompt a DNS lookup). This is what I took to be what is discussed as a "DNS Tickler" that others have stated cleared up DNS deafness as well as resolved the stuck email queue issue, for some at least. I have changed that though to use a lasso startup file following a post by Ke Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still hopeless in 8.6"
>
> CONCLUSION
>
> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.
>
> MONITORING
>
> There are three components to monitoring:
>
> 1) A lasso script that searches the site database and if errors are found then the display of the file is "restart" and if nothing found it is simply "ignore". This contains the logic and criteria for checking in the email queue databases for errors and stuck email. Under development, see the SQL at the bottom for a draft of the statement to search the database.
>
> 2) Commmand line scripts that perform the stop and start of Lasso. I do not see why not to use the pre-built scripts provided for in Lasso:
>
>    /Applications/Lasso Professional 8/Tools/startLassoService.command
>    /Applications/Lasso Professional 8/Tools/stopLassoService.command
>
> 3) Monitoring tool to load the URL in part 1 and if the result is "restart" then to use the scripts in 2 to stop then start again.
>
> For this third function, I am looking into using MONIT following the instructions at:
>
> "Monitoring Lasso with Monit"
> www.lassosoft.com/Monitoring-Lasso-with-Monit
>
> However, it seems that most of these instructions are outdated, for Lasso 8.5
>
> The only change I made on the monit configuration file "monitrc" is to add a line:
>
> include /etc/monit/*
>
> Then, I created a new sub-directory, named "monit" (within "etc"), for monit configuration files and made a new file within that named "lasso.monitrc"
>
> Here is where I need help. What to place in the file "lasso.monitrc". This is what I have so far:
>
> _________ start _________________
> check process lasso_0
>    with pidfile /tmp/LassoService8.pid
>    start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>        with timeout 120 seconds
>    stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>        with timeout 45 seconds
>
>    if changed pid then alert
>
>    if totalcpu > 15% for 2 cycles then alert
>    if totalmem > 75 MB for 2 cycles then alert
>
>    if 3 restarts within 5 cycles then timeout
>
>    group lasso
>    if totalcpu > 25% for 2 cycles then alert
>    if totalmem > 100 MB for 2 cycles then alert
>    depends on lasso_0
>
> check host myserver with address www.mydomain.com
>    stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>        with timeout 45 seconds
>    start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>        with timeout 120 seconds
>
>    if failed url <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso?display=none">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?display=none
>       and content == 'restart'
>       then restart
> _________ end _________________
>
> QUESTIONS / HELP
>
> Note that it seems that Lasso 8.6 is already created a .pid file for Lasso service, I did not use any additional tag as has been suggested. I only have one site and, for now at least, I am planning just to restart the Lasso Service. Perhaps I need though to have a pid file created for the one default site (only one default site on this web server)?
>
> Where can I get more details on how to create the monit configuration script?
>
> Maybe I should give up on the Monit monitoring and expand my lasso script to do the stop and start using the third-party os_process tag? Then schedule execution of that in Lasso Admin as an event??
>
> Michael
> __________________________________________________________________
>
> SELECT sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
> FROM smtp_queue AS sq \
> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND smtp_error.error LIKE "%Failed to connect to host%"
> ORDER BY sq.create_date DESC
>
> Michael Collins, Web Systems Engineer
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>
#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Marc Vos
1 Do you have any webpages scheduled in the EventQueue?
If so, it is really worth it to run each of them individually, directly in a webpage to see if one of them is having errors or generating any output - when in the queue, you don't see that, you don't know where it is routed to and thus these pages should not generate any output!

2 Another thing that helps is what Steve Piercy suggested in an earlier thread about the email-queue: disable error-output to LassoErrors.txt (switch on only when you need to, then switch it off again) and delete it from your sites folder, i.e. 'default-1'.

What Patrick does, does not work for me, for example. I also have a monitor (Lasso) running that sends an email out every 3 hours. But the error may still hit SiteAdmin's email-system, and then these mails do not get out either.
Since I went through the first two steps, I've had no more troubles with sending email.

- -
Marc



On 28 feb. 2013, at 05:40, Patrick Larkin <[hidden email]> wrote:

> I've seemed to have mostly resolved this issue by simply scheduling a lasso file that sends me an email every three hours.  I've only had the failure message you describe once in the past several months.  When I stop getting the alert email, I check the queue.  Not sure why this works but it seems to do the job for me.
>
> Patrick
>
> On Feb 27, 2013, at 9:22 PM, Michael <[hidden email]> wrote:
>
>> Hello all.
>>
>> Mac OS X server 10.6.8
>> Lasso 8.6
>>
>> I am struggling lately with the dreaded stuck email queue issue. The website that has been effected has not previously experienced this issue over the past several years. There have been no recent changes and the Website did not get hit with a spike in traffic and email load. However, in the last few weeks the email queue "Failed to connect to host" issue has surfaced. It's almost as if every email that is attempted uses the maximum retries until failure with the cannot connect to host error. The only way to get things going again is to either restart Lasso service fully, or restart the site and then re-queue all messages.
>>
>> ATTEMPTED SOLUTIONS
>>
>> I have tried:
>>
>> o Use IP numbers in place of domain names.
>>
>> o Use an alternate email server.
>>
>> o Send directly to SMTP Servers set to yes (but now back to No.)
>>
>> o Send an email every 30 minutes in some hope that would prevent services from going to a "sleep" state - which later did not seem to be the case.
>>
>> o Create an event that calls a lasso tag every 5 minutes (does an include_url to prompt a DNS lookup). This is what I took to be what is discussed as a "DNS Tickler" that others have stated cleared up DNS deafness as well as resolved the stuck email queue issue, for some at least. I have changed that though to use a lasso startup file following a post by Ke Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still hopeless in 8.6"
>>
>> CONCLUSION
>>
>> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.
>>
>> MONITORING
>>
>> There are three components to monitoring:
>>
>> 1) A lasso script that searches the site database and if errors are found then the display of the file is "restart" and if nothing found it is simply "ignore". This contains the logic and criteria for checking in the email queue databases for errors and stuck email. Under development, see the SQL at the bottom for a draft of the statement to search the database.
>>
>> 2) Commmand line scripts that perform the stop and start of Lasso. I do not see why not to use the pre-built scripts provided for in Lasso:
>>
>>   /Applications/Lasso Professional 8/Tools/startLassoService.command
>>   /Applications/Lasso Professional 8/Tools/stopLassoService.command
>>
>> 3) Monitoring tool to load the URL in part 1 and if the result is "restart" then to use the scripts in 2 to stop then start again.
>>
>> For this third function, I am looking into using MONIT following the instructions at:
>>
>> "Monitoring Lasso with Monit"
>> www.lassosoft.com/Monitoring-Lasso-with-Monit
>>
>> However, it seems that most of these instructions are outdated, for Lasso 8.5
>>
>> The only change I made on the monit configuration file "monitrc" is to add a line:
>>
>> include /etc/monit/*
>>
>> Then, I created a new sub-directory, named "monit" (within "etc"), for monit configuration files and made a new file within that named "lasso.monitrc"
>>
>> Here is where I need help. What to place in the file "lasso.monitrc". This is what I have so far:
>>
>> _________ start _________________
>> check process lasso_0
>>   with pidfile /tmp/LassoService8.pid
>>   start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>>       with timeout 120 seconds
>>   stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>>       with timeout 45 seconds
>>
>>   if changed pid then alert
>>
>>   if totalcpu > 15% for 2 cycles then alert
>>   if totalmem > 75 MB for 2 cycles then alert
>>
>>   if 3 restarts within 5 cycles then timeout
>>
>>   group lasso
>>   if totalcpu > 25% for 2 cycles then alert
>>   if totalmem > 100 MB for 2 cycles then alert
>>   depends on lasso_0
>>
>> check host myserver with address www.mydomain.com
>>   stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>>       with timeout 45 seconds
>>   start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>>       with timeout 120 seconds
>>
>>   if failed url <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso?display=none">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?display=none
>>      and content == 'restart'
>>      then restart
>> _________ end _________________
>>
>> QUESTIONS / HELP
>>
>> Note that it seems that Lasso 8.6 is already created a .pid file for Lasso service, I did not use any additional tag as has been suggested. I only have one site and, for now at least, I am planning just to restart the Lasso Service. Perhaps I need though to have a pid file created for the one default site (only one default site on this web server)?
>>
>> Where can I get more details on how to create the monit configuration script?
>>
>> Maybe I should give up on the Monit monitoring and expand my lasso script to do the stop and start using the third-party os_process tag? Then schedule execution of that in Lasso Admin as an event??
>>
>> Michael
>> __________________________________________________________________
>>
>> SELECT sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>> FROM smtp_queue AS sq \
>> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND smtp_error.error LIKE "%Failed to connect to host%"
>> ORDER BY sq.create_date DESC
>>
>> Michael Collins, Web Systems Engineer
>>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

stevepiercy
On 2/28/13 at 9:29 AM, [hidden email] (Marc Vos) pronounced:

>1 Do you have any webpages scheduled in the EventQueue? If so,
>it is really worth it to run each of them individually,
>directly in a webpage to see if one of them is having errors or
>generating any output - when in the queue, you don't see that,
>you don't know where it is routed to and thus these pages
>should not generate any output!
>
>2 Another thing that helps is what Steve Piercy suggested in an
>earlier thread about the email-queue: disable error-output to
>LassoErrors.txt (switch on only when you need to, then switch
>it off again) and delete it from your sites folder, i.e. 'default-1'.

In general, error logging in production should be minimized and
turned on only when absolutely necessary.  I've seen silly
things, like retaining successful messages enabled in email
setup.  Keep your logs and databases rotated and clean to
minimize issues with opening and writing to bloated files.

--steve


>What Patrick does, does not work for me, for example. I also
>have a monitor (Lasso) running that sends an email out every 3
>hours. But the error may still hit SiteAdmin's email-system,
>and then these mails do not get out either.
>Since I went through the first two steps, I've had no more
>troubles with sending email.
>
>- -
>Marc
>
>
>
>On 28 feb. 2013, at 05:40, Patrick Larkin <[hidden email]> wrote:
>
>>I've seemed to have mostly resolved this issue by simply scheduling a lasso file
>that sends me an email every three hours.  I've only had the
>failure message you describe once in the past several months.  
>When I stop getting the alert email, I check the queue.  Not
>sure why this works but it seems to do the job for me.
>>
>>Patrick
>>
>>On Feb 27, 2013, at 9:22 PM, Michael <[hidden email]> wrote:
>>
>>>  Hello all.   Mac OS X server 10.6.8
>>> Lasso 8.6
>>>   I am struggling lately with the dreaded stuck email queue
>>>issue. The website
>that has been effected has not previously experienced this
>issue over the past several years. There have been no recent
>changes and the Website did not get hit with a spike in traffic
>and email load. However, in the last few weeks the email queue
>"Failed to connect to host" issue has surfaced. It's almost as
>if every email that is attempted uses the maximum retries until
>failure with the cannot connect to host error. The only way to
>get things going again is to either restart Lasso service
>fully, or restart the site and then re-queue all messages.
>>>   ATTEMPTED SOLUTIONS
>>>   I have tried:
>>>   o Use IP numbers in place of domain names.
>>>   o Use an alternate email server.
>>>   o Send directly to SMTP Servers set to yes (but now back
>>>to No.)   o Send an email every 30 minutes in some hope that
>>>would prevent services from
>going to a "sleep" state - which later did not seem to be the case.
>>>   o Create an event that calls a lasso tag every 5 minutes
>>>(does an include_url
>to prompt a DNS lookup). This is what I took to be what is
>discussed as a "DNS Tickler" that others have stated cleared up
>DNS deafness as well as resolved the stuck email queue issue,
>for some at least. I have changed that though to use a lasso
>startup file following a post by Ke Carlton on Dec 06, 2011;
>8:11pm in the thread "Email system still hopeless in 8.6"
>>>   CONCLUSION
>>>   What I have decided now is that the current email queue
>>>system CANNOT BE FIXED
>and but that I must live with monitoring for a stuck email
>queue and restarting Lasso automatically when that problem is found.
>>>   MONITORING
>>>   There are three components to monitoring:
>>>   1) A lasso script that searches the site database and if
>>>errors are found then
>the display of the file is "restart" and if nothing found it is
>simply "ignore". This contains the logic and criteria for
>checking in the email queue databases for errors and stuck
>email. Under development, see the SQL at the bottom for a draft
>of the statement to search the database.
>>>   2) Commmand line scripts that perform the stop and start
>>>of Lasso. I do not see
>why not to use the pre-built scripts provided for in Lasso:
>>>     /Applications/Lasso Professional 8/Tools/startLassoService.command
>>>   /Applications/Lasso Professional 8/Tools/stopLassoService.command
>>>   3) Monitoring tool to load the URL in part 1 and if the
>>>result is "restart"
>then to use the scripts in 2 to stop then start again.
>>>   For this third function, I am looking into using MONIT
>>>following the
>instructions at:
>>>   "Monitoring Lasso with Monit"  www.lassosoft.com/Monitoring-Lasso-with-Monit
>>>   However, it seems that most of these instructions are
>>>outdated, for Lasso 8.5
>>>   The only change I made on the monit configuration file
>>>"monitrc" is to add a
>line:
>>>   include /etc/monit/*
>>>   Then, I created a new sub-directory, named "monit" (within
>>>"etc"), for monit
>configuration files and made a new file within that named "lasso.monitrc"
>>>   Here is where I need help. What to place in the file
>>>"lasso.monitrc". This is
>what I have so far:
>>>   _________ start _________________
>>> check process lasso_0
>>>   with pidfile /tmp/LassoService8.pid
>>>   start program = "/Applications/Lasso Professional
>8/Tools/startLassoService.command"
>>>       with timeout 120 seconds
>>>   stop program = "/Applications/Lasso Professional
>8/Tools/stopLassoService.command"
>>>       with timeout 45 seconds
>>>     if changed pid then alert
>>>     if totalcpu > 15% for 2 cycles then alert
>>>   if totalmem > 75 MB for 2 cycles then alert
>>>     if 3 restarts within 5 cycles then timeout
>>>     group lasso
>>>   if totalcpu > 25% for 2 cycles then alert
>>>   if totalmem > 100 MB for 2 cycles then alert
>>>   depends on lasso_0
>>>   check host myserver with address www.mydomain.com
>>>   stop program = "/Applications/Lasso Professional
>8/Tools/stopLassoService.command"
>>>       with timeout 45 seconds
>>>   start program = "/Applications/Lasso Professional
>8/Tools/startLassoService.command"
>>>       with timeout 120 seconds
>>>     if failed url
><a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?
>display=none
>>>      and content == 'restart'
>>>      then restart
>>> _________ end _________________
>>>   QUESTIONS / HELP
>>>   Note that it seems that Lasso 8.6 is already created a
>>>.pid file for Lasso
>service, I did not use any additional tag as has been
>suggested. I only have one site and, for now at least, I am
>planning just to restart the Lasso Service. Perhaps I need
>though to have a pid file created for the one default site
>(only one default site on this web server)?
>>>   Where can I get more details on how to create the monit
>>>configuration script?
>>>   Maybe I should give up on the Monit monitoring and expand
>>>my lasso script to do
>the stop and start using the third-party os_process tag? Then
>schedule execution of that in Lasso Admin as an event??
>>>   Michael
>>> __________________________________________________________________
>>>   SELECT
>sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.
>smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.
>smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.
>queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>>> FROM smtp_queue AS sq \
>>> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>>> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR
>sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL
>AND smtp_error.error LIKE "%Failed to connect to host%"
>>> ORDER BY sq.create_date DESC
>>>   Michael Collins, Web Systems Engineer
>>>
>
>#############################################################
>This message is sent to you because you are subscribed to
>the mailing list Lasso
>[hidden email]
>To unsubscribe, E-mail to: <[hidden email]>
>Send administrative queries to  <[hidden email]>

-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
-- --
Steve Piercy               Web Site Builder              
Soquel, CA
<[hidden email]>                  <http://www.StevePiercy.com/>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

mail
In reply to this post by Patrick Larkin-2
Patrick

Thanks for the input, however, I did see your post and attempted that fix. It did not work for my situtation.

Michael


On Feb 28, 2013, at 12:40 PM, Patrick Larkin wrote:

>
> I've seemed to have mostly resolved this issue by simply scheduling a lasso file that sends me an email every three hours.  I've only had the failure message you describe once in the past several months.  When I stop getting the alert email, I check the queue.  Not sure why this works but it seems to do the job for me.
>
> Patrick
>
> On Feb 27, 2013, at 9:22 PM, Michael <[hidden email]> wrote:
>
>> Hello all.
>>
>> Mac OS X server 10.6.8
>> Lasso 8.6
>>
>> I am struggling lately with the dreaded stuck email queue issue. The website that has been effected has not previously experienced this issue over the past several years. There have been no recent changes and the Website did not get hit with a spike in traffic and email load. However, in the last few weeks the email queue "Failed to connect to host" issue has surfaced. It's almost as if every email that is attempted uses the maximum retries until failure with the cannot connect to host error. The only way to get things going again is to either restart Lasso service fully, or restart the site and then re-queue all messages.
>>
>> ATTEMPTED SOLUTIONS
>>
>> I have tried:
>>
>> o Use IP numbers in place of domain names.
>>
>> o Use an alternate email server.
>>
>> o Send directly to SMTP Servers set to yes (but now back to No.)
>>
>> o Send an email every 30 minutes in some hope that would prevent services from going to a "sleep" state - which later did not seem to be the case.
>>
>> o Create an event that calls a lasso tag every 5 minutes (does an include_url to prompt a DNS lookup). This is what I took to be what is discussed as a "DNS Tickler" that others have stated cleared up DNS deafness as well as resolved the stuck email queue issue, for some at least. I have changed that though to use a lasso startup file following a post by Ke Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still hopeless in 8.6"
>>
>> CONCLUSION
>>
>> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.
>>
>> MONITORING
>>
>> There are three components to monitoring:
>>
>> 1) A lasso script that searches the site database and if errors are found then the display of the file is "restart" and if nothing found it is simply "ignore". This contains the logic and criteria for checking in the email queue databases for errors and stuck email. Under development, see the SQL at the bottom for a draft of the statement to search the database.
>>
>> 2) Commmand line scripts that perform the stop and start of Lasso. I do not see why not to use the pre-built scripts provided for in Lasso:
>>
>>   /Applications/Lasso Professional 8/Tools/startLassoService.command
>>   /Applications/Lasso Professional 8/Tools/stopLassoService.command
>>
>> 3) Monitoring tool to load the URL in part 1 and if the result is "restart" then to use the scripts in 2 to stop then start again.
>>
>> For this third function, I am looking into using MONIT following the instructions at:
>>
>> "Monitoring Lasso with Monit"
>> www.lassosoft.com/Monitoring-Lasso-with-Monit
>>
>> However, it seems that most of these instructions are outdated, for Lasso 8.5
>>
>> The only change I made on the monit configuration file "monitrc" is to add a line:
>>
>> include /etc/monit/*
>>
>> Then, I created a new sub-directory, named "monit" (within "etc"), for monit configuration files and made a new file within that named "lasso.monitrc"
>>
>> Here is where I need help. What to place in the file "lasso.monitrc". This is what I have so far:
>>
>> _________ start _________________
>> check process lasso_0
>>   with pidfile /tmp/LassoService8.pid
>>   start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>>       with timeout 120 seconds
>>   stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>>       with timeout 45 seconds
>>
>>   if changed pid then alert
>>
>>   if totalcpu > 15% for 2 cycles then alert
>>   if totalmem > 75 MB for 2 cycles then alert
>>
>>   if 3 restarts within 5 cycles then timeout
>>
>>   group lasso
>>   if totalcpu > 25% for 2 cycles then alert
>>   if totalmem > 100 MB for 2 cycles then alert
>>   depends on lasso_0
>>
>> check host myserver with address www.mydomain.com
>>   stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>>       with timeout 45 seconds
>>   start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>>       with timeout 120 seconds
>>
>>   if failed url <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso?display=none">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?display=none
>>      and content == 'restart'
>>      then restart
>> _________ end _________________
>>
>> QUESTIONS / HELP
>>
>> Note that it seems that Lasso 8.6 is already created a .pid file for Lasso service, I did not use any additional tag as has been suggested. I only have one site and, for now at least, I am planning just to restart the Lasso Service. Perhaps I need though to have a pid file created for the one default site (only one default site on this web server)?
>>
>> Where can I get more details on how to create the monit configuration script?
>>
>> Maybe I should give up on the Monit monitoring and expand my lasso script to do the stop and start using the third-party os_process tag? Then schedule execution of that in Lasso Admin as an event??
>>
>> Michael
>> __________________________________________________________________
>>
>> SELECT sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>> FROM smtp_queue AS sq \
>> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND smtp_error.error LIKE "%Failed to connect to host%"
>> ORDER BY sq.create_date DESC
>>
>> Michael Collins, Web Systems Engineer
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>
>

Michael Collins, Web Systems Engineer

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Seth Ganahl
Michael-

I can only say that I'm struggling with the same issue myself, and have been
for almost a year now.

Lasso 8.6.0
Mac OS X Server 10.6

I restart the Lasso Service, and requeue all errors in the SMTP queue
probably 5-10 times a day.

I've been told by LassoSoft that upgrading to Lasso 9 is the fix, so I'm
sitting tight until I can get the client to do that.


On 2/28/13 5:40 AM, "Michael" <[hidden email]> did quoth:

> Patrick
>
> Thanks for the input, however, I did see your post and attempted that fix. It
> did not work for my situtation.
>
> Michael
>
>
> On Feb 28, 2013, at 12:40 PM, Patrick Larkin wrote:
>
>>
>> I've seemed to have mostly resolved this issue by simply scheduling a lasso
>> file that sends me an email every three hours.  I've only had the failure
>> message you describe once in the past several months.  When I stop getting
>> the alert email, I check the queue.  Not sure why this works but it seems to
>> do the job for me.
>>
>> Patrick
>>
>> On Feb 27, 2013, at 9:22 PM, Michael <[hidden email]> wrote:
>>
>>> Hello all.
>>>
>>> Mac OS X server 10.6.8
>>> Lasso 8.6
>>>
>>> I am struggling lately with the dreaded stuck email queue issue. The website
>>> that has been effected has not previously experienced this issue over the
>>> past several years. There have been no recent changes and the Website did
>>> not get hit with a spike in traffic and email load. However, in the last few
>>> weeks the email queue "Failed to connect to host" issue has surfaced. It's
>>> almost as if every email that is attempted uses the maximum retries until
>>> failure with the cannot connect to host error. The only way to get things
>>> going again is to either restart Lasso service fully, or restart the site
>>> and then re-queue all messages.
>>>
>>> ATTEMPTED SOLUTIONS
>>>
>>> I have tried:
>>>
>>> o Use IP numbers in place of domain names.
>>>
>>> o Use an alternate email server.
>>>
>>> o Send directly to SMTP Servers set to yes (but now back to No.)
>>>
>>> o Send an email every 30 minutes in some hope that would prevent services
>>> from going to a "sleep" state - which later did not seem to be the case.
>>>
>>> o Create an event that calls a lasso tag every 5 minutes (does an
>>> include_url to prompt a DNS lookup). This is what I took to be what is
>>> discussed as a "DNS Tickler" that others have stated cleared up DNS deafness
>>> as well as resolved the stuck email queue issue, for some at least. I have
>>> changed that though to use a lasso startup file following a post by Ke
>>> Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still hopeless
>>> in 8.6"
>>>
>>> CONCLUSION
>>>
>>> What I have decided now is that the current email queue system CANNOT BE
>>> FIXED and but that I must live with monitoring for a stuck email queue and
>>> restarting Lasso automatically when that problem is found.
>>>
>>> MONITORING
>>>
>>> There are three components to monitoring:
>>>
>>> 1) A lasso script that searches the site database and if errors are found
>>> then the display of the file is "restart" and if nothing found it is simply
>>> "ignore". This contains the logic and criteria for checking in the email
>>> queue databases for errors and stuck email. Under development, see the SQL
>>> at the bottom for a draft of the statement to search the database.
>>>
>>> 2) Commmand line scripts that perform the stop and start of Lasso. I do not
>>> see why not to use the pre-built scripts provided for in Lasso:
>>>
>>>   /Applications/Lasso Professional 8/Tools/startLassoService.command
>>>   /Applications/Lasso Professional 8/Tools/stopLassoService.command
>>>
>>> 3) Monitoring tool to load the URL in part 1 and if the result is "restart"
>>> then to use the scripts in 2 to stop then start again.
>>>
>>> For this third function, I am looking into using MONIT following the
>>> instructions at:
>>>
>>> "Monitoring Lasso with Monit"
>>> www.lassosoft.com/Monitoring-Lasso-with-Monit
>>>
>>> However, it seems that most of these instructions are outdated, for Lasso
>>> 8.5
>>>
>>> The only change I made on the monit configuration file "monitrc" is to add a
>>> line:
>>>
>>> include /etc/monit/*
>>>
>>> Then, I created a new sub-directory, named "monit" (within "etc"), for monit
>>> configuration files and made a new file within that named "lasso.monitrc"
>>>
>>> Here is where I need help. What to place in the file "lasso.monitrc". This
>>> is what I have so far:
>>>
>>> _________ start _________________
>>> check process lasso_0
>>>   with pidfile /tmp/LassoService8.pid
>>>   start program = "/Applications/Lasso Professional
>>> 8/Tools/startLassoService.command"
>>>       with timeout 120 seconds
>>>   stop program = "/Applications/Lasso Professional
>>> 8/Tools/stopLassoService.command"
>>>       with timeout 45 seconds
>>>
>>>   if changed pid then alert
>>>
>>>   if totalcpu > 15% for 2 cycles then alert
>>>   if totalmem > 75 MB for 2 cycles then alert
>>>
>>>   if 3 restarts within 5 cycles then timeout
>>>
>>>   group lasso
>>>   if totalcpu > 25% for 2 cycles then alert
>>>   if totalmem > 100 MB for 2 cycles then alert
>>>   depends on lasso_0
>>>
>>> check host myserver with address www.mydomain.com
>>>   stop program = "/Applications/Lasso Professional
>>> 8/Tools/stopLassoService.command"
>>>       with timeout 45 seconds
>>>   start program = "/Applications/Lasso Professional
>>> 8/Tools/startLassoService.command"
>>>       with timeout 120 seconds
>>>
>>>   if failed url
>>> <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.la">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.la
>>> sso?display=none
>>>      and content == 'restart'
>>>      then restart
>>> _________ end _________________
>>>
>>> QUESTIONS / HELP
>>>
>>> Note that it seems that Lasso 8.6 is already created a .pid file for Lasso
>>> service, I did not use any additional tag as has been suggested. I only have
>>> one site and, for now at least, I am planning just to restart the Lasso
>>> Service. Perhaps I need though to have a pid file created for the one
>>> default site (only one default site on this web server)?
>>>
>>> Where can I get more details on how to create the monit configuration
>>> script?
>>>
>>> Maybe I should give up on the Monit monitoring and expand my lasso script to
>>> do the stop and start using the third-party os_process tag? Then schedule
>>> execution of that in Lasso Admin as an event??
>>>
>>> Michael
>>> __________________________________________________________________
>>>
>>> SELECT
>>> sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.
>>> smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.
>>> smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.que
>>> ue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>>> FROM smtp_queue AS sq \
>>> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>>> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR
>>> sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND
>>> smtp_error.error LIKE "%Failed to connect to host%"
>>> ORDER BY sq.create_date DESC
>>>
>>> Michael Collins, Web Systems Engineer
>>>
>>> #############################################################
>>> This message is sent to you because you are subscribed to
>>> the mailing list Lasso
>>> [hidden email]
>>> To unsubscribe, E-mail to: <[hidden email]>
>>> Send administrative queries to  <[hidden email]>
>> #############################################################
>> This message is sent to you because you are subscribed to
>>  the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>>
>
> Michael Collins, Web Systems Engineer
>
> #############################################################
> This message is sent to you because you are subscribed to
>   the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>

-------------------------------------------------------
Seth C Ganahl (501) 282-4867
Ganahl Consulting ­ Web Applications
http://www.ganahlconsulting.com/
[hidden email]
-------------------------------------------------------


#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Anibal Escobar
Has anyone tried sending messages via an external mail server using  
the smtp tags (email_smtp, smtp->open, smtp->send, smtp->close)?  I  
use this with a client who sends out thousands of emails and I don't  
want them going through my mail server. I believe these messages  
bypass the lasso mail queue as the smtp tags allow you to communicate  
directly with the mail server.  I do this in Lasso 8.5.6.

Thanks,
Anibal Escobar
646 336 5000 x8821




On Mar 1, 2013, at 9:58 AM, Seth Ganahl wrote:

> Michael-
>
> I can only say that I'm struggling with the same issue myself, and  
> have been
> for almost a year now.
>
> Lasso 8.6.0
> Mac OS X Server 10.6
>
> I restart the Lasso Service, and requeue all errors in the SMTP queue
> probably 5-10 times a day.
>
> I've been told by LassoSoft that upgrading to Lasso 9 is the fix, so  
> I'm
> sitting tight until I can get the client to do that.
>
>
> On 2/28/13 5:40 AM, "Michael" <[hidden email]> did quoth:
>
>> Patrick
>>
>> Thanks for the input, however, I did see your post and attempted  
>> that fix. It
>> did not work for my situtation.
>>
>> Michael
>>
>>
>> On Feb 28, 2013, at 12:40 PM, Patrick Larkin wrote:
>>
>>>
>>> I've seemed to have mostly resolved this issue by simply  
>>> scheduling a lasso
>>> file that sends me an email every three hours.  I've only had the  
>>> failure
>>> message you describe once in the past several months.  When I stop  
>>> getting
>>> the alert email, I check the queue.  Not sure why this works but  
>>> it seems to
>>> do the job for me.
>>>
>>> Patrick
>>>
>>> On Feb 27, 2013, at 9:22 PM, Michael <[hidden email]>  
>>> wrote:
>>>
>>>> Hello all.
>>>>
>>>> Mac OS X server 10.6.8
>>>> Lasso 8.6
>>>>
>>>> I am struggling lately with the dreaded stuck email queue issue.  
>>>> The website
>>>> that has been effected has not previously experienced this issue  
>>>> over the
>>>> past several years. There have been no recent changes and the  
>>>> Website did
>>>> not get hit with a spike in traffic and email load. However, in  
>>>> the last few
>>>> weeks the email queue "Failed to connect to host" issue has  
>>>> surfaced. It's
>>>> almost as if every email that is attempted uses the maximum  
>>>> retries until
>>>> failure with the cannot connect to host error. The only way to  
>>>> get things
>>>> going again is to either restart Lasso service fully, or restart  
>>>> the site
>>>> and then re-queue all messages.
>>>>
>>>> ATTEMPTED SOLUTIONS
>>>>
>>>> I have tried:
>>>>
>>>> o Use IP numbers in place of domain names.
>>>>
>>>> o Use an alternate email server.
>>>>
>>>> o Send directly to SMTP Servers set to yes (but now back to No.)
>>>>
>>>> o Send an email every 30 minutes in some hope that would prevent  
>>>> services
>>>> from going to a "sleep" state - which later did not seem to be  
>>>> the case.
>>>>
>>>> o Create an event that calls a lasso tag every 5 minutes (does an
>>>> include_url to prompt a DNS lookup). This is what I took to be  
>>>> what is
>>>> discussed as a "DNS Tickler" that others have stated cleared up  
>>>> DNS deafness
>>>> as well as resolved the stuck email queue issue, for some at  
>>>> least. I have
>>>> changed that though to use a lasso startup file following a post  
>>>> by Ke
>>>> Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still  
>>>> hopeless
>>>> in 8.6"
>>>>
>>>> CONCLUSION
>>>>
>>>> What I have decided now is that the current email queue system  
>>>> CANNOT BE
>>>> FIXED and but that I must live with monitoring for a stuck email  
>>>> queue and
>>>> restarting Lasso automatically when that problem is found.
>>>>
>>>> MONITORING
>>>>
>>>> There are three components to monitoring:
>>>>
>>>> 1) A lasso script that searches the site database and if errors  
>>>> are found
>>>> then the display of the file is "restart" and if nothing found it  
>>>> is simply
>>>> "ignore". This contains the logic and criteria for checking in  
>>>> the email
>>>> queue databases for errors and stuck email. Under development,  
>>>> see the SQL
>>>> at the bottom for a draft of the statement to search the database.
>>>>
>>>> 2) Commmand line scripts that perform the stop and start of  
>>>> Lasso. I do not
>>>> see why not to use the pre-built scripts provided for in Lasso:
>>>>
>>>>  /Applications/Lasso Professional 8/Tools/startLassoService.command
>>>>  /Applications/Lasso Professional 8/Tools/stopLassoService.command
>>>>
>>>> 3) Monitoring tool to load the URL in part 1 and if the result is  
>>>> "restart"
>>>> then to use the scripts in 2 to stop then start again.
>>>>
>>>> For this third function, I am looking into using MONIT following  
>>>> the
>>>> instructions at:
>>>>
>>>> "Monitoring Lasso with Monit"
>>>> www.lassosoft.com/Monitoring-Lasso-with-Monit
>>>>
>>>> However, it seems that most of these instructions are outdated,  
>>>> for Lasso
>>>> 8.5
>>>>
>>>> The only change I made on the monit configuration file "monitrc"  
>>>> is to add a
>>>> line:
>>>>
>>>> include /etc/monit/*
>>>>
>>>> Then, I created a new sub-directory, named "monit" (within  
>>>> "etc"), for monit
>>>> configuration files and made a new file within that named  
>>>> "lasso.monitrc"
>>>>
>>>> Here is where I need help. What to place in the file  
>>>> "lasso.monitrc". This
>>>> is what I have so far:
>>>>
>>>> _________ start _________________
>>>> check process lasso_0
>>>>  with pidfile /tmp/LassoService8.pid
>>>>  start program = "/Applications/Lasso Professional
>>>> 8/Tools/startLassoService.command"
>>>>      with timeout 120 seconds
>>>>  stop program = "/Applications/Lasso Professional
>>>> 8/Tools/stopLassoService.command"
>>>>      with timeout 45 seconds
>>>>
>>>>  if changed pid then alert
>>>>
>>>>  if totalcpu > 15% for 2 cycles then alert
>>>>  if totalmem > 75 MB for 2 cycles then alert
>>>>
>>>>  if 3 restarts within 5 cycles then timeout
>>>>
>>>>  group lasso
>>>>  if totalcpu > 25% for 2 cycles then alert
>>>>  if totalmem > 100 MB for 2 cycles then alert
>>>>  depends on lasso_0
>>>>
>>>> check host myserver with address www.mydomain.com
>>>>  stop program = "/Applications/Lasso Professional
>>>> 8/Tools/stopLassoService.command"
>>>>      with timeout 45 seconds
>>>>  start program = "/Applications/Lasso Professional
>>>> 8/Tools/startLassoService.command"
>>>>      with timeout 120 seconds
>>>>
>>>>  if failed url
>>>> <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.la">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.la
>>>> sso?display=none
>>>>     and content == 'restart'
>>>>     then restart
>>>> _________ end _________________
>>>>
>>>> QUESTIONS / HELP
>>>>
>>>> Note that it seems that Lasso 8.6 is already created a .pid file  
>>>> for Lasso
>>>> service, I did not use any additional tag as has been suggested.  
>>>> I only have
>>>> one site and, for now at least, I am planning just to restart the  
>>>> Lasso
>>>> Service. Perhaps I need though to have a pid file created for the  
>>>> one
>>>> default site (only one default site on this web server)?
>>>>
>>>> Where can I get more details on how to create the monit  
>>>> configuration
>>>> script?
>>>>
>>>> Maybe I should give up on the Monit monitoring and expand my  
>>>> lasso script to
>>>> do the stop and start using the third-party os_process tag? Then  
>>>> schedule
>>>> execution of that in Lasso Admin as an event??
>>>>
>>>> Michael
>>>> __________________________________________________________________
>>>>
>>>> SELECT
>>>> sq
>>>> .id
>>>> ,sq
>>>> .id_site
>>>> ,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.
>>>> smtp_host
>>>> ,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.
>>>> smtp_ssl
>>>> ,sq
>>>> .queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.que
>>>> ue_priority
>>>> ,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>>>> FROM smtp_queue AS sq \
>>>> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>>>> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE  
>>>> "%sending%" OR
>>>> sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND
>>>> smtp_error.error LIKE "%Failed to connect to host%"
>>>> ORDER BY sq.create_date DESC
>>>>
>>>> Michael Collins, Web Systems Engineer
>>>>
>>>> #############################################################
>>>> This message is sent to you because you are subscribed to
>>>> the mailing list Lasso
>>>> [hidden email]
>>>> To unsubscribe, E-mail to: <[hidden email]>
>>>> Send administrative queries to  <[hidden email]>
>>> #############################################################
>>> This message is sent to you because you are subscribed to
>>> the mailing list Lasso
>>> [hidden email]
>>> To unsubscribe, E-mail to: <[hidden email]>
>>> Send administrative queries to  <[hidden email]>
>>>
>>
>> Michael Collins, Web Systems Engineer
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>>  the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
> -------------------------------------------------------
> Seth C Ganahl (501) 282-4867
> Ganahl Consulting – Web Applications
> http://www.ganahlconsulting.com/
> [hidden email]
> -------------------------------------------------------
>
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

mail
In reply to this post by stevepiercy
When I downloaded the RD_lassoService_pid tag and attempted to use it, it would not load in LassoStartup. I did not realize it immediately until I attempted to run it in a browser (surrounding it with an inline containing permissions to use OS_Process). However, I got the error:

Lasso8Service(39099,0xb0081000) malloc: *** mmap(size=1342177280) failed (error code=12)
*** error: can't allocate region
*** set a breakpoint in malloc_error_break to debug
-1 UNKNOWN EXCEPTION while running startup script "".

I then went through RD_lassoService_pid.inc and cleaned it up. Basically, I rewrote the page on a new file because it appeared to be some sort of corruption. I am not sure exactly which changed fixed it. One thing I did was change the file to .lasso. Took out all the comments. Took out the extra [ ] square brackets that surrounded the page. I also moved the calling of the RD_lassoService_pid tag WITHIN the  if(lasso_tagexists('os_process')) conditional. But, it works now.

Michael

On Feb 28, 2013, at 11:12 AM, Steve Piercy - Web Site Builder wrote:

>
> Regarding:
> http://www.lassosoft.com/Monitoring-Lasso-with-Monit
>
> See:
> http://lasso.2283332.n4.nabble.com/Monit-with-Lasso-8-6-on-CentOS-tp4635649p4635654.html
>
> Note the differences between Mac and CentOS, and 8.5 and 8.6 regarding the PIDs.  Marc Pinnell made comments under [RD_lassoService_pid] that are relevant to your observation:
> http://www.lassosoft.com/tagswap/detail/RD_lassoService_pid
>
> If you use the [RD_lassoService_pid] tag as recommended, then you merely restart the Lasso instance, not server, and reduce restart time.
>
> --steve
>
>
> On 2/28/13 at 10:22 AM, [hidden email] (Michael) pronounced:
>
>> Hello all.
>> Mac OS X server 10.6.8
>> Lasso 8.6
>>
>> I am struggling lately with the dreaded stuck email queue issue. The website that has been effected has not previously experienced this issue over the past several years. There have been no recent changes and the Website did not get hit with a spike in traffic and email load. However, in the last few weeks the email queue "Failed to connect to host" issue has surfaced. It's almost as if every email that is attempted uses the maximum retries until failure with the cannot connect to host error. The only way to get things going again is to either restart Lasso service fully, or restart the site and then re-queue all messages.
>> ATTEMPTED SOLUTIONS
>>
>> I have tried:
>>
>> o Use IP numbers in place of domain names.
>>
>> o Use an alternate email server.
>>
>> o Send directly to SMTP Servers set to yes (but now back to No.)
>> o Send an email every 30 minutes in some hope that would prevent services from going to a "sleep" state - which later did not seem to be the case.
>>
>> o Create an event that calls a lasso tag every 5 minutes (does an include_url to prompt a DNS lookup). This is what I took to be what is discussed as a "DNS Tickler" that others have stated cleared up DNS deafness as well as resolved the stuck email queue issue, for some at least. I have changed that though to use a lasso startup file following a post by Ke Carlton on Dec 06, 2011; 8:11pm in the thread "Email system still hopeless in 8.6"
>>
>> CONCLUSION
>>
>> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.
>> MONITORING
>>
>> There are three components to monitoring:
>>
>> 1) A lasso script that searches the site database and if errors are found then the display of the file is "restart" and if nothing found it is simply "ignore". This contains the logic and criteria for checking in the email queue databases for errors and stuck email. Under development, see the SQL at the bottom for a draft of the statement to search the database.
>>
>> 2) Commmand line scripts that perform the stop and start of Lasso. I do not see why not to use the pre-built scripts provided for in Lasso:
>> /Applications/Lasso Professional 8/Tools/startLassoService.command
>> /Applications/Lasso Professional 8/Tools/stopLassoService.command
>>
>> 3) Monitoring tool to load the URL in part 1 and if the result is "restart" then to use the scripts in 2 to stop then start again.
>> For this third function, I am looking into using MONIT following the instructions at:
>>
>> "Monitoring Lasso with Monit" www.lassosoft.com/Monitoring-Lasso-with-Monit
>>
>> However, it seems that most of these instructions are outdated, for Lasso 8.5
>>
>> The only change I made on the monit configuration file "monitrc" is to add a line:
>>
>> include /etc/monit/*
>>
>> Then, I created a new sub-directory, named "monit" (within "etc"), for monit configuration files and made a new file within that named "lasso.monitrc"
>>
>> Here is where I need help. What to place in the file "lasso.monitrc". This is what I have so far:
>>
>> _________ start _________________
>> check process lasso_0
>> with pidfile /tmp/LassoService8.pid
>> start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>> with timeout 120 seconds
>> stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>> with timeout 45 seconds
>>
>> if changed pid then alert
>>
>> if totalcpu > 15% for 2 cycles then alert
>> if totalmem > 75 MB for 2 cycles then alert
>>
>> if 3 restarts within 5 cycles then timeout
>>
>> group lasso
>> if totalcpu > 25% for 2 cycles then alert
>> if totalmem > 100 MB for 2 cycles then alert
>> depends on lasso_0
>>
>> check host myserver with address www.mydomain.com
>> stop program = "/Applications/Lasso Professional 8/Tools/stopLassoService.command"
>> with timeout 45 seconds
>> start program = "/Applications/Lasso Professional 8/Tools/startLassoService.command"
>> with timeout 120 seconds
>>
>> if failed url <a href="http://robotuser:wL8%l@neDBY7h2B490@www.mydomain.com/checkemailqueueerror.lasso">http://robotuser:wL8%l@neDBY7h2B490@.../checkemailqueueerror.lasso?
>> display=none
>> and content == 'restart'
>> then restart
>> _________ end _________________
>>
>> QUESTIONS / HELP
>>
>> Note that it seems that Lasso 8.6 is already created a .pid file for Lasso service, I did not use any additional tag as has been suggested. I only have one site and, for now at least, I am planning just to restart the Lasso Service. Perhaps I need though to have a pid file created for the one default site (only one default site on this web server)?
>>
>> Where can I get more details on how to create the monit configuration script?
>>
>> Maybe I should give up on the Monit monitoring and expand my lasso script to do the stop and start using the third-party os_process tag? Then schedule execution of that in Lasso Admin as an event??
>>
>> Michael
>> __________________________________________________________________
>>
>> SELECT sq.id,sq.id_site,sq.id_route,sq.id_cache,sq.mime_recipients,sq.mime_from,sq.
>> smtp_host,sq.smtp_port,sq.smtp_timeout,sq.smtp_username,sq.smtp_password,sq.
>> smtp_ssl,sq.queue_date,sq.queue_status,sq.queue_error,sq.queue_number,sq.
>> queue_priority,sq.queue_retries,sq.merge_array,sq.merge_tokens,sq.create_date \
>> FROM smtp_queue AS sq \
>> LEFT OUTER JOIN smtp_error ON(sq.id=smtp_error.id_email) \
>> WHERE (sq.queue_status LIKE "%error%" OR sq.queue_status LIKE "%sending%" OR sq.queue_status LIKE "%queued%") AND smtp_error.id IS NOT NULL AND smtp_error.error LIKE "%Failed to connect to host%"
>> ORDER BY sq.create_date DESC
>>
>> Michael Collins, Web Systems Engineer
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
> -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
> Steve Piercy               Web Site Builder               Soquel, CA
> <[hidden email]>                  <http://www.StevePiercy.com/>
>
> #############################################################
> This message is sent to you because you are subscribed to
> the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>
>

Michael Collins, Web Systems Engineer

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Erik Vandermey - WebCentrix, Inc.
In reply to this post by mail
My situation:

Mac OS X Server 10.8.2
Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)

I've tried all the voodoo mentioned before. I have nothing to add other than I need a solution that does not necessitate babysitting the Lasso Email Queue.

I may try creating my own email sending routine as someone has mentioned before, but it would be nice if LassoSoft could fix this bug.

I have two LassoSites where email notifications are CRITICAL.


Regards,
---
Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers  
Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller



On Feb 27, 2013, at 8:22 PM, Michael <[hidden email]> wrote:

> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.


#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Anibal Escobar
I haven't tested this as a custom tag, but I use a variant of the code  
below to send emails bypassing the lasso queue.

define_tag('send', -nameSpace='emailAlt_', -required='host', -
required='port', -required='from', -required='to', -
required='subject', -required='body');
        local('SMTP'=Email_SMTP);
        #SMTP->Open(-Host=#host,-port=#port);
        if(error_currenterror(-errorcode) != 0);
                error code here
        /if;
        local('msg'= Email_compose(-from=#from, -to=#to, -subject=#subject, -
body=#body));
        local('ok' = #SMTP->Send(-From=#msg->From, -Recipients=#msg-
 >Recipients, -Message=#msg->Data))
        If(#ok);
                success code if any
        Else;
                error code here
        /if;
        #SMTP->Close;
/define;

This requires that the mail server you use as #host allow your machine  
to relay messages through it.

Thanks,
Anibal Escobar
646 336 5000 x8821




On Mar 8, 2013, at 2:48 PM, Erik Vandermey - WebCentrix, Inc. wrote:

> My situation:
>
> Mac OS X Server 10.8.2
> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>
> I've tried all the voodoo mentioned before. I have nothing to add  
> other than I need a solution that does not necessitate babysitting  
> the Lasso Email Queue.
>
> I may try creating my own email sending routine as someone has  
> mentioned before, but it would be nice if LassoSoft could fix this  
> bug.
>
> I have two LassoSites where email notifications are CRITICAL.
>
>
> Regards,
> ---
> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/
> Dedicated Servers
> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator  
> Reseller
>
>
>
> On Feb 27, 2013, at 8:22 PM, Michael <[hidden email]> wrote:
>
>> What I have decided now is that the current email queue system  
>> CANNOT BE FIXED and but that I must live with monitoring for a  
>> stuck email queue and restarting Lasso automatically when that  
>> problem is found.
>
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Erik Vandermey - WebCentrix, Inc.
Mac OS X Server 10.8.2
Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)

I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.

So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...

On Mar 8, 2013, at 2:46 PM, Anibal Escobar <[hidden email]> wrote:

> I haven't tested this as a custom tag, but I use a variant of the code below to send emails bypassing the lasso queue.
>
> define_tag('send', -nameSpace='emailAlt_', -required='host', -required='port', -required='from', -required='to', -required='subject', -required='body');
> local('SMTP'=Email_SMTP);
> #SMTP->Open(-Host=#host,-port=#port);
> if(error_currenterror(-errorcode) != 0);
> error code here
> /if;
> local('msg'= Email_compose(-from=#from, -to=#to, -subject=#subject, -body=#body));
> local('ok' = #SMTP->Send(-From=#msg->From, -Recipients=#msg->Recipients, -Message=#msg->Data))
> If(#ok);
> success code if any
> Else;
> error code here
> /if;
> #SMTP->Close;
> /define;
>
> This requires that the mail server you use as #host allow your machine to relay messages through it.
>
> Thanks,
> Anibal Escobar
> 646 336 5000 x8821
>
>
>
>
> On Mar 8, 2013, at 2:48 PM, Erik Vandermey - WebCentrix, Inc. wrote:
>
>> My situation:
>>
>> Mac OS X Server 10.8.2
>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>
>> I've tried all the voodoo mentioned before. I have nothing to add other than I need a solution that does not necessitate babysitting the Lasso Email Queue.
>>
>> I may try creating my own email sending routine as someone has mentioned before, but it would be nice if LassoSoft could fix this bug.
>>
>> I have two LassoSites where email notifications are CRITICAL.
>>
>>
>> Regards,
>> ---
>> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
>> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers
>> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller
>>
>>
>>
>> On Feb 27, 2013, at 8:22 PM, Michael <[hidden email]> wrote:
>>
>>> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.
>>
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
> #############################################################
> This message is sent to you because you are subscribed to
> the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>


#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Marc Pope-2
Tried using OpenDNS before? I have much better luck with them than my upstream providers DNS servers.

On Mar 13, 2013, at 6:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:

> Mac OS X Server 10.8.2
> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>
> I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.
>
> So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...
>
> On Mar 8, 2013, at 2:46 PM, Anibal Escobar <[hidden email]> wrote:
>
>> I haven't tested this as a custom tag, but I use a variant of the code below to send emails bypassing the lasso queue.
>>
>> define_tag('send', -nameSpace='emailAlt_', -required='host', -required='port', -required='from', -required='to', -required='subject', -required='body');
>>    local('SMTP'=Email_SMTP);
>>    #SMTP->Open(-Host=#host,-port=#port);
>>    if(error_currenterror(-errorcode) != 0);
>>        error code here
>>    /if;
>>    local('msg'= Email_compose(-from=#from, -to=#to, -subject=#subject, -body=#body));
>>    local('ok' = #SMTP->Send(-From=#msg->From, -Recipients=#msg->Recipients, -Message=#msg->Data))
>>    If(#ok);
>>        success code if any
>>    Else;
>>        error code here
>>    /if;
>>    #SMTP->Close;
>> /define;
>>
>> This requires that the mail server you use as #host allow your machine to relay messages through it.
>>
>> Thanks,
>> Anibal Escobar
>> 646 336 5000 x8821
>>
>>
>>
>>
>> On Mar 8, 2013, at 2:48 PM, Erik Vandermey - WebCentrix, Inc. wrote:
>>
>>> My situation:
>>>
>>> Mac OS X Server 10.8.2
>>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>>
>>> I've tried all the voodoo mentioned before. I have nothing to add other than I need a solution that does not necessitate babysitting the Lasso Email Queue.
>>>
>>> I may try creating my own email sending routine as someone has mentioned before, but it would be nice if LassoSoft could fix this bug.
>>>
>>> I have two LassoSites where email notifications are CRITICAL.
>>>
>>>
>>> Regards,
>>> ---
>>> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
>>> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers
>>> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller
>>>
>>>
>>>
>>> On Feb 27, 2013, at 8:22 PM, Michael <[hidden email]> wrote:
>>>
>>>> What I have decided now is that the current email queue system CANNOT BE FIXED and but that I must live with monitoring for a stuck email queue and restarting Lasso automatically when that problem is found.
>>>
>>>
>>> #############################################################
>>> This message is sent to you because you are subscribed to
>>> the mailing list Lasso
>>> [hidden email]
>>> To unsubscribe, E-mail to: <[hidden email]>
>>> Send administrative queries to  <[hidden email]>
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>
#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

fletcher sandbeck-2
In reply to this post by Erik Vandermey - WebCentrix, Inc.
On Mar 13, 2013, at 3:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:

> Mac OS X Server 10.8.2
> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>
> I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.
>
> So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...

This is really why the email queue exists in the first place.  You can use [Email_Send] with an -Immediate tag and the SMTP interaction will occur immediately, but if there is a problem on your server or with the remote SMTP server then the email will be lost.  The queue ensures that the email is stored so that once the server problems are sorted out it can be sent.

What I'd like to see is an option to take advantage of either a local relay like sendmail or a third-party API like those which Elastic Email or other transactional email services provide.  That would speed up the mail queue and offload the actual work of SMTP interaction.  Plus, services like Elastic return an ID that can be used for further API calls so you could get information about deliverability and open rates.

[fletcher]

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Erik Vandermey - WebCentrix, Inc.
Already sending/relaying to a local email server also behind the firewall. Configured the SMTP_Host option with an IP address instead of a domain name. Tried various external DNS servers (e.g. Google DNS - 8.8.8.8) and local DNS server behind the firewall. When DNS lookups fail, email doesn't send.

I don't know if there is a way to tell Email_Send or Email_SMTP to ignore a DNS lookup when an IP address is used for the host. They may not be programmed that way.

Regards,
---
Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers  
Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller



On Mar 13, 2013, at 5:59 PM, Fletcher Sandbeck <[hidden email]> wrote:

> On Mar 13, 2013, at 3:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:
>
>> Mac OS X Server 10.8.2
>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>
>> I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.
>>
>> So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...
>
> This is really why the email queue exists in the first place.  You can use [Email_Send] with an -Immediate tag and the SMTP interaction will occur immediately, but if there is a problem on your server or with the remote SMTP server then the email will be lost.  The queue ensures that the email is stored so that once the server problems are sorted out it can be sent.
>
> What I'd like to see is an option to take advantage of either a local relay like sendmail or a third-party API like those which Elastic Email or other transactional email services provide.  That would speed up the mail queue and offload the actual work of SMTP interaction.  Plus, services like Elastic return an ID that can be used for further API calls so you could get information about deliverability and open rates.
>
> [fletcher]
>
> #############################################################
> This message is sent to you because you are subscribed to
> the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>


#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Marc Pope-2
FYI I have a mail server on the same server and I use localhost as my nail server in lasso.. And not send direct with lasso, I have a lot better success that way.

Marc

On Mar 13, 2013, at 7:12 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:

> Already sending/relaying to a local email server also behind the firewall. Configured the SMTP_Host option with an IP address instead of a domain name. Tried various external DNS servers (e.g. Google DNS - 8.8.8.8) and local DNS server behind the firewall. When DNS lookups fail, email doesn't send.
>
> I don't know if there is a way to tell Email_Send or Email_SMTP to ignore a DNS lookup when an IP address is used for the host. They may not be programmed that way.
>
> Regards,
> ---
> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers  
> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller
>
>
>
> On Mar 13, 2013, at 5:59 PM, Fletcher Sandbeck <[hidden email]> wrote:
>
>> On Mar 13, 2013, at 3:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:
>>
>>> Mac OS X Server 10.8.2
>>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>>
>>> I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.
>>>
>>> So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...
>>
>> This is really why the email queue exists in the first place.  You can use [Email_Send] with an -Immediate tag and the SMTP interaction will occur immediately, but if there is a problem on your server or with the remote SMTP server then the email will be lost.  The queue ensures that the email is stored so that once the server problems are sorted out it can be sent.
>>
>> What I'd like to see is an option to take advantage of either a local relay like sendmail or a third-party API like those which Elastic Email or other transactional email services provide.  That would speed up the mail queue and offload the actual work of SMTP interaction.  Plus, services like Elastic return an ID that can be used for further API calls so you could get information about deliverability and open rates.
>>
>> [fletcher]
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>
#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Erik Vandermey - WebCentrix, Inc.
OK, tried setting up DNS and email on same server as Lasso (running 10.8.2 server). Ran into the same problem. As soon as DNS lookups fail, all email stops working (either email_send or email_smtp).

Configured DNS to use 127.0.0.1 and email to use 127.0.0.1. Setup local users to authenticate, relay all email thru our regular mail server (also on same IP subnet). Works for a little while than BAM no more DNS lookups (took less than a 1/2 hour for DNS to quit working).

I'm almost at a point where I'm going to stuff email messages into a MySQL table and have PHP pull them out and send them via a shell script.

How sad is that.


On Mar 13, 2013, at 6:15 PM, Marc Pope <[hidden email]> wrote:

> FYI I have a mail server on the same server and I use localhost as my nail server in lasso.. And not send direct with lasso, I have a lot better success that way.
>
> Marc
>
> On Mar 13, 2013, at 7:12 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:
>
>> Already sending/relaying to a local email server also behind the firewall. Configured the SMTP_Host option with an IP address instead of a domain name. Tried various external DNS servers (e.g. Google DNS - 8.8.8.8) and local DNS server behind the firewall. When DNS lookups fail, email doesn't send.
>>
>> I don't know if there is a way to tell Email_Send or Email_SMTP to ignore a DNS lookup when an IP address is used for the host. They may not be programmed that way.
>>
>> Regards,
>> ---
>> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
>> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers  
>> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller
>>
>>
>>
>> On Mar 13, 2013, at 5:59 PM, Fletcher Sandbeck <[hidden email]> wrote:
>>
>>> On Mar 13, 2013, at 3:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:
>>>
>>>> Mac OS X Server 10.8.2
>>>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>>>
>>>> I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.
>>>>
>>>> So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...
>>>
>>> This is really why the email queue exists in the first place.  You can use [Email_Send] with an -Immediate tag and the SMTP interaction will occur immediately, but if there is a problem on your server or with the remote SMTP server then the email will be lost.  The queue ensures that the email is stored so that once the server problems are sorted out it can be sent.
>>>
>>> What I'd like to see is an option to take advantage of either a local relay like sendmail or a third-party API like those which Elastic Email or other transactional email services provide.  That would speed up the mail queue and offload the actual work of SMTP interaction.  Plus, services like Elastic return an ID that can be used for further API calls so you could get information about deliverability and open rates.
>>>
>>> [fletcher]
>>>
>>> #############################################################
>>> This message is sent to you because you are subscribed to
>>> the mailing list Lasso
>>> [hidden email]
>>> To unsubscribe, E-mail to: <[hidden email]>
>>> Send administrative queries to  <[hidden email]>
>>
>>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
> #############################################################
> This message is sent to you because you are subscribed to
> the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>


#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Anibal Escobar
So I guess the root of the problem is the DNS lookup failing.  How do  
you know that the DNS lookup fails?  Is the DNS lookup you're  
referring to Lasso trying to find the mail server, or the mail server  
trying to find the MX records of the domains to which the messages are  
going?

Thanks,
Anibal Escobar
646 336 5000 x8821




On Mar 13, 2013, at 10:23 PM, Erik Vandermey - WebCentrix, Inc. wrote:

> OK, tried setting up DNS and email on same server as Lasso (running  
> 10.8.2 server). Ran into the same problem. As soon as DNS lookups  
> fail, all email stops working (either email_send or email_smtp).
>
> Configured DNS to use 127.0.0.1 and email to use 127.0.0.1. Setup  
> local users to authenticate, relay all email thru our regular mail  
> server (also on same IP subnet). Works for a little while than BAM  
> no more DNS lookups (took less than a 1/2 hour for DNS to quit  
> working).
>
> I'm almost at a point where I'm going to stuff email messages into a  
> MySQL table and have PHP pull them out and send them via a shell  
> script.
>
> How sad is that.
>
>
> On Mar 13, 2013, at 6:15 PM, Marc Pope <[hidden email]> wrote:
>
>> FYI I have a mail server on the same server and I use localhost as  
>> my nail server in lasso.. And not send direct with lasso, I have a  
>> lot better success that way.
>>
>> Marc
>>
>> On Mar 13, 2013, at 7:12 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]
>> > wrote:
>>
>>> Already sending/relaying to a local email server also behind the  
>>> firewall. Configured the SMTP_Host option with an IP address  
>>> instead of a domain name. Tried various external DNS servers (e.g.  
>>> Google DNS - 8.8.8.8) and local DNS server behind the firewall.  
>>> When DNS lookups fail, email doesn't send.
>>>
>>> I don't know if there is a way to tell Email_Send or Email_SMTP to  
>>> ignore a DNS lookup when an IP address is used for the host. They  
>>> may not be programmed that way.
>>>
>>> Regards,
>>> ---
>>> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
>>> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/
>>> Dedicated Servers
>>> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator  
>>> Reseller
>>>
>>>
>>>
>>> On Mar 13, 2013, at 5:59 PM, Fletcher Sandbeck  
>>> <[hidden email]> wrote:
>>>
>>>> On Mar 13, 2013, at 3:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]
>>>> > wrote:
>>>>
>>>>> Mac OS X Server 10.8.2
>>>>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>>>>
>>>>> I rewrote some email sending routines using the example below,  
>>>>> and included adding the email message parameters to a SQL table  
>>>>> just in case the email_SMTP tag errors out. Good thing I did,  
>>>>> Email_SMTP apparently suffers from the same problem as  
>>>>> Email_Send in that when DNS goes deaf, the email won't send in  
>>>>> either case until I restart Lasso for that site.
>>>>>
>>>>> So although the below code results in more efficient sending, it  
>>>>> only works as long as DNS doesn't go deaf. Filing a bug report...
>>>>
>>>> This is really why the email queue exists in the first place.  
>>>> You can use [Email_Send] with an -Immediate tag and the SMTP  
>>>> interaction will occur immediately, but if there is a problem on  
>>>> your server or with the remote SMTP server then the email will be  
>>>> lost.  The queue ensures that the email is stored so that once  
>>>> the server problems are sorted out it can be sent.
>>>>
>>>> What I'd like to see is an option to take advantage of either a  
>>>> local relay like sendmail or a third-party API like those which  
>>>> Elastic Email or other transactional email services provide.  
>>>> That would speed up the mail queue and offload the actual work of  
>>>> SMTP interaction.  Plus, services like Elastic return an ID that  
>>>> can be used for further API calls so you could get information  
>>>> about deliverability and open rates.
>>>>
>>>> [fletcher]
>>>>
>>>> #############################################################
>>>> This message is sent to you because you are subscribed to
>>>> the mailing list Lasso
>>>> [hidden email]
>>>> To unsubscribe, E-mail to: <[hidden email]>
>>>> Send administrative queries to  <[hidden email]>
>>>
>>>
>>> #############################################################
>>> This message is sent to you because you are subscribed to
>>> the mailing list Lasso
>>> [hidden email]
>>> To unsubscribe, E-mail to: <[hidden email]>
>>> Send administrative queries to  <[hidden email]>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Marc Pope-2
In reply to this post by Erik Vandermey - WebCentrix, Inc.
Have you tried 'localhost' instead of 127.0.0.1 ?

And DNS use: 208.67.222.222 and 208.67.220.220 which is open DNS. Make sure to restart lasso.

Marc


On Mar 13, 2013, at 10:23 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:

> OK, tried setting up DNS and email on same server as Lasso (running 10.8.2 server). Ran into the same problem. As soon as DNS lookups fail, all email stops working (either email_send or email_smtp).
>
> Configured DNS to use 127.0.0.1 and email to use 127.0.0.1. Setup local users to authenticate, relay all email thru our regular mail server (also on same IP subnet). Works for a little while than BAM no more DNS lookups (took less than a 1/2 hour for DNS to quit working).
>
> I'm almost at a point where I'm going to stuff email messages into a MySQL table and have PHP pull them out and send them via a shell script.
>
> How sad is that.
>
>
> On Mar 13, 2013, at 6:15 PM, Marc Pope <[hidden email]> wrote:
>
>> FYI I have a mail server on the same server and I use localhost as my nail server in lasso.. And not send direct with lasso, I have a lot better success that way.
>>
>> Marc
>>
>> On Mar 13, 2013, at 7:12 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:
>>
>>> Already sending/relaying to a local email server also behind the firewall. Configured the SMTP_Host option with an IP address instead of a domain name. Tried various external DNS servers (e.g. Google DNS - 8.8.8.8) and local DNS server behind the firewall. When DNS lookups fail, email doesn't send.
>>>
>>> I don't know if there is a way to tell Email_Send or Email_SMTP to ignore a DNS lookup when an IP address is used for the host. They may not be programmed that way.
>>>
>>> Regards,
>>> ---
>>> Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
>>> WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers  
>>> Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller
>>>
>>>
>>>
>>> On Mar 13, 2013, at 5:59 PM, Fletcher Sandbeck <[hidden email]> wrote:
>>>
>>>> On Mar 13, 2013, at 3:50 PM, "Erik Vandermey - WebCentrix, Inc." <[hidden email]> wrote:
>>>>
>>>>> Mac OS X Server 10.8.2
>>>>> Lasso 8.6.0.1 (reports as 8.6.0 in the admin console)
>>>>>
>>>>> I rewrote some email sending routines using the example below, and included adding the email message parameters to a SQL table just in case the email_SMTP tag errors out. Good thing I did, Email_SMTP apparently suffers from the same problem as Email_Send in that when DNS goes deaf, the email won't send in either case until I restart Lasso for that site.
>>>>>
>>>>> So although the below code results in more efficient sending, it only works as long as DNS doesn't go deaf. Filing a bug report...
>>>>
>>>> This is really why the email queue exists in the first place.  You can use [Email_Send] with an -Immediate tag and the SMTP interaction will occur immediately, but if there is a problem on your server or with the remote SMTP server then the email will be lost.  The queue ensures that the email is stored so that once the server problems are sorted out it can be sent.
>>>>
>>>> What I'd like to see is an option to take advantage of either a local relay like sendmail or a third-party API like those which Elastic Email or other transactional email services provide.  That would speed up the mail queue and offload the actual work of SMTP interaction.  Plus, services like Elastic return an ID that can be used for further API calls so you could get information about deliverability and open rates.
>>>>
>>>> [fletcher]
>>>>
>>>> #############################################################
>>>> This message is sent to you because you are subscribed to
>>>> the mailing list Lasso
>>>> [hidden email]
>>>> To unsubscribe, E-mail to: <[hidden email]>
>>>> Send administrative queries to  <[hidden email]>
>>>
>>>
>>> #############################################################
>>> This message is sent to you because you are subscribed to
>>> the mailing list Lasso
>>> [hidden email]
>>> To unsubscribe, E-mail to: <[hidden email]>
>>> Send administrative queries to  <[hidden email]>
>> #############################################################
>> This message is sent to you because you are subscribed to
>> the mailing list Lasso
>> [hidden email]
>> To unsubscribe, E-mail to: <[hidden email]>
>> Send administrative queries to  <[hidden email]>
>
>
> #############################################################
> This message is sent to you because you are subscribed to
>  the mailing list Lasso
> [hidden email]
> To unsubscribe, E-mail to: <[hidden email]>
> Send administrative queries to  <[hidden email]>
#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: stuck email queue - seeking automated lasso restart

Erik Vandermey - WebCentrix, Inc.
In reply to this post by Anibal Escobar
All DNS lookups fail. I verify this by going to the siteadmin.lassoapp for the given site experiencing problems:

Setup > Site > Lasso Settings screen and click on the "Test" button in the "Domain Name Lookups" section.

Regards,
---
Erik Van-der-Mey • [hidden email] • (AIM/iChat: [hidden email])
WebCentrix, Inc. • http://webcentrix.net • Web Hosting/Co-Location/Dedicated Servers  
Certified Lasso Developer (CLD) • Kerio Connect/Control/Operator Reseller



On Mar 14, 2013, at 10:10 AM, Anibal Escobar <[hidden email]> wrote:

> So I guess the root of the problem is the DNS lookup failing.  How do you know that the DNS lookup fails?  Is the DNS lookup you're referring to Lasso trying to find the mail server, or the mail server trying to find the MX records of the domains to which the messages are going?


#############################################################
This message is sent to you because you are subscribed to
  the mailing list Lasso
[hidden email]
To unsubscribe, E-mail to: <[hidden email]>
Send administrative queries to  <[hidden email]>
123