Pound server get too many CLOSE_WAIT sessions and fails
Bug #1189764 reported by
Daniel Bidwell
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
pound (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
We upgraded to the 3.2.0-45-generic kernel on a pair of 64 but vms and now the server gets lots of errors with connectoins with CLOSED_WAIT status. This eventually fills up the connection tables and new work stops happening. We have removed any iptables firewalls from the pound servers and the lvs load balancer in front of them. I have also back revved the kernels back to 3.2.0-24-generic, but it doesn't look like ths is helping.
These servers are all running ubuntu 12.04.02 for x64. They have been running in this configuration for years without a problem until yesterday. Looking for new ideas to try to resolve this.
affects: | nginx (Ubuntu) → pound (Ubuntu) |
To post a comment you must log in.
This bug was fixed in the package pound - 3.0-1
---------------
pound (3.0-1) unstable; urgency=medium
[ Carsten Leonhardt ]
* New upstream version
As pound 3.0 was rewritten from scratch, also most Ubuntu bug reports
no longer apply (LP: #1189764, LP: #1398007, LP: #1639888)
* Upstream is out of beta, upload to unstable
* Add debian/NEWS to inform about new configuration format and file name
[ Debian Janitor ] upstream/ metadata (already
* debian/copyright: use spaces rather than tabs to start continuation lines.
* Remove obsolete fields Contact, Name from debian/
present in machine-readable debian/copyright).
* Fix day-of-week for changelog entries 2.5-1.1, 2.4.3-1, 2.4.2-1, 2.4-2.
-- Carsten Leonhardt <email address hidden> Thu, 05 Nov 2020 01:45:01 +0100