- attached corrupted_email_2020_SEP_11.zip
Delivery queue in loop, high CPU and memory usage
Hi,
Sometimes my server goes on 100% CPU and high memory usage, and evenually it become irresponsible.
I’ve learned that threre are many messages stuck in the folder /queue/smtp
and /queue/submission
, all with the same error message appended in every attempt, in loop.
The solution is to manually delete that folder every time the processor reaches more than 90% usage
I’m using Version 2.2.22 PRO # 949
Thats a piece of the message, the full content is in attachment.
Let me know If I could provide more logs or details about the problem.
"attachment_archive_files": [],
"data_line_length_exceeded": true,
"authentication_results": [
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
"dkim=fail (body hash did not verify) header.i=@multiclubes.com.br",
... more thousand lines ...
Comments (6)
-
reporter -
reporter - changed status to open
-
reporter It still happens, even after upgrading to the latest release (2.2.22 - 20200901).
I have deleted all queued messages again, just to confirm it.
-
reporter There is a new information …
The recpt_to is teste3@teste.com.br
That domain “teste.com.br” resolves to 127.0.0.1 (not in my DNS, but globally)
I think that is the loop cause
-
reporter One more case: mail sent to @gamail.com (yes gmail spelled wrong)
the mx.gamail.com reports as localhost.localdomain, and causes a loop too
The workarround was to create a virtual domain bin for both “teste.com.br” and “gamail.com” so the loop ends
-
repo owner - changed status to resolved
Try latest version, there was multiple internal changes which might solve the issue. If it doesn't help please reopen
- Log in to comment
New sample files that I just deleted from
/queue/smtp