OpenVZ Forum


Home » General » Discussions » Nslcd and SSHd die on 2.6.32-5-openvz-amd64 debian
Nslcd and SSHd die on 2.6.32-5-openvz-amd64 debian [message #45592] Wed, 21 March 2012 11:29
spo0nman is currently offline  spo0nman
Messages: 1
Registered: March 2012
Junior Member
From: *119.232.166.static-Bangalore.vsnl.net.in
I've been observing this for a few days now. On the Openvz containers. The nslcd deamon dies with the following error:

Mar 12 11:27:03 erp-test1 nslcd[6290]: [8b4567] ldap_result() failed: Size limit exceeded
Mar 12 11:27:06 erp-test1 nslcd[6290]: [7b23c6] ldap_result() failed: Size limit exceeded
Mar 12 11:42:03 erp-test1 nslcd[6290]: [221a70] error writing to client: Broken pipe
Mar 12 11:42:03 erp-test1 nslcd[6290]: [465f01] error writing to client: Broken pipe
Mar 12 11:47:26 erp-test1 nslcd[6290]: [305def] error writing to client: Broken pipe

I ran the daemon in debug mode and the error I saw was:

nslcd: [1f671a] DEBUG: ldap_unbind()
nslcd: [74368c] DEBUG: ldap_unbind()
nslcd: [85782f] DEBUG: ldap_unbind()
nslcd: [74f9a9] DEBUG: ldap_unbind()
nslcd: [b14ba2] DEBUG: ldap_unbind()
nslcd: caught signal SIGHUP (1), shutting down
nslcd: version 0.7.15 bailing out

At the same time sshd also dies for some reason -- I've not debugged further. It could also be that sshd dies taking nslcd with it.

This has happened on multiple containers on multiple hardware configurations. This only happens to containers and not the master host.

I also noticed that when I send a kill -HUP to nslcd it dies.

I'm running Linux test1 2.6.32-5-openvz-amd64 #1 SMP Mon Oct 3 05:12:50 UTC 2011 x86_64 GNU/Linux

Help?
Previous Topic: vmstat FPE
Next Topic: CentOS 6 x64 Minimal Template
Goto Forum:
  


Current Time: Sun Jul 12 22:16:49 GMT 2020