Bug 745105

Summary: File descriptors warning on re-synchronization
Product: [Retired] 389 Reporter: Diego Woitasen <diego>
Component: Directory ServerAssignee: Rich Megginson <rmeggins>
Status: CLOSED INSUFFICIENT_DATA QA Contact: Chandrasekar Kannan <ckannan>
Severity: low Docs Contact:
Priority: unspecified    
Version: 1.2.9CC: benl
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-10-20 21:31:45 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Diego Woitasen 2011-10-11 12:09:39 UTC
Description of problem:
When I do a replication re-sync I get a "Config Warning" about file descriptors. This message only appears on re-sync, DS doesn't complain about FDs when I restart the service. The message is:

[11/Oct/2011:06:21:01 -0300] - All database threads now stopped
[11/Oct/2011:06:21:01 -0300] - Config Warning: - nsslapd-maxdescriptors: invalid value "10240", maximum file descriptors must range from 1 to 1024 (the current process limit).  Server will use a setting of 1024.
[11/Oct/2011:06:21:01 -0300] - Config Warning: - nsslapd-reservedescriptors: invalid value "5275", reserved file descriptors must range from 1 to 1024 (the current process maxdescriptors limit).  Server will use a setting of 1024.
[11/Oct/2011:06:21:09 -0300] - export shared: Processed 17 entries (100%).
[11/Oct/2011:06:21:09 -0300] - All database threads now stopped

I run ulimit -n 10240 in the init script. It looks like the new process that re-rsync the DBs isn't taking that limit.


Version-Release number of selected component (if applicable):
Tested with 1.2.8.3 and 1.2.9.9

Comment 1 Rich Megginson 2011-10-12 14:48:47 UTC
(In reply to comment #0)
> Description of problem:
> When I do a replication re-sync I get a "Config Warning" about file
> descriptors. This message only appears on re-sync, DS doesn't complain about
> FDs when I restart the service. The message is:
> 
> [11/Oct/2011:06:21:01 -0300] - All database threads now stopped
> [11/Oct/2011:06:21:01 -0300] - Config Warning: - nsslapd-maxdescriptors:
> invalid value "10240", maximum file descriptors must range from 1 to 1024 (the
> current process limit).  Server will use a setting of 1024.
> [11/Oct/2011:06:21:01 -0300] - Config Warning: - nsslapd-reservedescriptors:
> invalid value "5275", reserved file descriptors must range from 1 to 1024 (the
> current process maxdescriptors limit).  Server will use a setting of 1024.
> [11/Oct/2011:06:21:09 -0300] - export shared: Processed 17 entries (100%).
> [11/Oct/2011:06:21:09 -0300] - All database threads now stopped
> 
> I run ulimit -n 10240 in the init script. It looks like the new process that
> re-rsync the DBs isn't taking that limit.

Do you also change the number of file descriptors allowed in the kernel?
http://directory.fedoraproject.org/wiki/Performance_Tuning#Linux

> 
> 
> Version-Release number of selected component (if applicable):
> Tested with 1.2.8.3 and 1.2.9.9

Comment 2 Diego Woitasen 2011-10-12 14:53:02 UTC
Yes. I followed the recommendations of that link when I setup the servers. The things is that the ns-slapd doesn't log anything about FD limit when you start the service. The message only appears on re-sync.

Comment 3 Rich Megginson 2011-10-12 15:04:21 UTC
(In reply to comment #2)
> Yes. I followed the recommendations of that link when I setup the servers. The
> things is that the ns-slapd doesn't log anything about FD limit when you start
> the service. The message only appears on re-sync.

By "re-sync" do you mean you do an ldapmodify operation to change the nsds5BeginReplicaRefresh: start on the supplier?

Comment 4 Diego Woitasen 2011-10-20 21:31:45 UTC
Sorry, I'm not sure about this bug now. It looks like the Warning appears when you dump the database with db2ldif.

I'll close it.