Mail archiving count is very low.

[Rajesh Samineni]

Dear JSuto,

This is to inform you that, we are facing the the issue while archiving maximum mails to archive, we are observing the maximum mails in queue to archive in mail server. I'm getting this issue after migrating the PILER setup to new server, Kindly suggest to us to tune the PILER to accept the maximum mail.

--Thanks, Rajesh Samineni, rajesh.saminei (skype).

[Janos SUTO]

Er, I'm not sure if I understand the issue. What are 'maximum mails'? Are they huge emails piler cannot accept? Please elaborate the issue.

[Rajesh Samineni]

Hi,

Hourly we recevies around 3000-4000 mails, same amount (all mails) of mails we are trying to archive.

Issue, but Piler accepting hundreds(300 -400) of mails out of thousands, remaining mails are showing in queue/ timed out in my mail server level.

[Janos SUTO]

Please check your maillog whether piler is actually receiving emails. If so then try to flush emails. Make sure the mail server doesn't try to deliver more emails parallel then the number of child processes in piler.

[Rajesh Samineni]

Yes, mails server is trying to deliver to piler, but getting timed out due to maximum mails. Let me know do I need to change my configuration. Currently we configured like worker_processed =10000 and max requests per child=25000.

[Janos SUTO]

Please explain what's maximum mail.

[Rajesh Samineni]

Here maximum mails mean current queue in mail server to deliver to piler, now it's around 100000.

[Janos SUTO]

Any errors in the mail server log?

[Rajesh Samineni]

Getting timed out error in mail server log

[Janos SUTO]

max_request_per_child=25000 may be ok, but number_of_worker_processes=10000 means that piler should fork 10000 child processes. I don't think you really want this. Normally 10-20 forked processes are just fine.

[Janos SUTO]

Any update on the matter?

[Janos SUTO]

As I suggested, you should match the number of 'threads' the mta sends to piler, and the num_of_worker_processes option in piler.conf.