I think the directives are similar for event and worker and any config changes
would not be too bad, but the worker documentation would give better info. To
switch I think you would load the worker module instead of event or recompile
with worker depending on how you are doing it.
Date: Tue, 2 Jun 2015 00:36:41 +0000
From: mark_jacq...@yahoo.com.INVALID
To: users@httpd.apache.org
Subject: Re: [users@httpd] AH00485: scoreboard is full, not at MaxRequestWorkers
I read that as well but the following statement he made did not make me feel
like it totally resolved the issue.
"However, what I did notice is that the same server periodically had all of its
workers taken up with requests, so that may be relevant to the problem as well."
If I did switch to "Worker" I wonder how much config file change would be
needed?
Any thoughts on that?
MJ
On Monday, June 1, 2015 5:33 PM, Scot Russell <scotrussell...@hotmail.com>
wrote:
I don't know much on this, but for the bug link you included, one comment
mentioned it being a problem with event mpm, that went away when using worker
mpm. Is it possible to try the worker mpm?Date: Mon, 1 Jun 2015 23:58:13
+0000From: mark_jacq...@yahoo.com.INVALIDTo: users@httpd.apache.orgSubject:
[users@httpd] AH00485: scoreboard is full, not at MaxRequestWorkersSorry for
the bandwidth, this is a repeat send from January.We STILL have this issue.
Perhaps a new look might jolt your memory and help me resolve this
issue!!!Problem: Apache
server will stay up for random amount of time, usually days, but
eventually enters a hung state. When hung the CPU load gradually spikes
on the machine and new web server requests are unresponsive. Only a full server
stop/kill and restart will resolve the issue, at which time the host load
average drops quickly to normal.Error logs typically contain many many lines
of: Wed Jan 28 16:06:58.667188 2015] [mpm_event:error] [pid 25336:tid 1]
AH00485: scoreboard is full, not at MaxRequestWorkersI
have done a lot of web research on this top and have found many cases
where others have had the same/similar issue but no real solutions.
Seem very close to this bug report:
https://issues.apache.org/bugzilla/show_bug.cgi?id=53555Environment:LDOM (VM)
SunOS myhostname 5.10 Generic_118833-36 sun4v sparc SUNW,Sun-Fire-T200 8G
RAMhttp Conf:StartServers 8MinSpareServers
Not setMaxSpareServers Not setServerLimit
256MaxRequestWorkers 200MaxConnectionsPerChild 200KeepAlive
OnTimeOut 3000MaxKeepAliveRequests
50KeepAliveTimeout 2RequestReadTimeout header=10-20,MinRate=500
body=20,MinRate=500We use LDAP to authenitcate most of our web tools, and this
is configured like:## Enable the LDAP connection pool and shared# memory cache.
Enable the LDAP cache status# handler. Requires that mod_ldap and
mod_authnz_ldap be loaded.#LDAPSharedCacheSize 500000LDAPCacheEntries
1024LDAPCacheTTL 600LDAPOpCacheEntries 1024LDAPOpCacheTTL 600Current non-hung
Score Board:Apache Server Status for myserver.domain (via 5.5.5.5)
Server Version: Apache/2.4.12 (Unix)Server MPM: eventServer Built: Feb 20 2015
13:47:55
Current Time: Monday, 01-Jun-2015 16:52:14 PDTRestart Time: Monday, 01-Jun-2015
00:35:23 PDTParent Server Config. Generation: 1Parent Server MPM Generation:
0Server uptime: 16 hours 16 minutes 51 secondsServer load: 1.06 0.84 0.60Total
accesses: 8263 - Total Traffic: 414.7 MBCPU Usage: u20.9 s32 cu660.69 cs121.41
- 1.42% CPU load.141 requests/sec - 7.2 kB/second - 51.4 kB/request2 requests
currently being processed, 98 idle workers
PIDConnectionsThreadsAsync
connectionstotalacceptingbusyidlewritingkeep-aliveclosing187930yes02500095253yes124003134840yes025000195063yes124002Sum6
298005
........................._______________________________________
________W__________________________________W_________________...
................................................................
........Any thoughts/comments on http conf tuning, OS patches, apache bug fixes
appreciated.This is a production server, so you can imagine, having it go down
at random times (usually when I am asleep) is not fun!Thanks.MJ
