[SOLVED] zmcontrol stop; zmcontrol start needed several times per day

[dvandenputte]

Hello

new installation, new customer.


During the test phase ==> everything was OK.
Just after production: the server need to be restarted (zmcontrol stop; zmcontrol start), several times per day.

The customer reports (no login possible either using webmail or IMAP client).

The log reports (actual data masked)

May 16 17:07:39 dc1 postfix/lmtp[22480]: 3EBD01766D2E: to=<USER@DOMAIN>, relay=server.domain.com [10.67.129.1]:7025, delay=5206, delays=4906/0.02/300/0, dsn=4.4.2, status=deferred (conversation with server.domain.com[10.67.129.1] timed out while receiving the initial server greeting)

and (more obscure to me)

May 16 15:50:46 dc1 saslauthd[9885]: zmpost: url='https://server.domain.com:7071/service/admin/soap/' returned buffer->data='<soap:Envelope xmlns:soap="http://www.w3.org/2003/05/soap-envelope"><soap:Header><context xmlns="urn:zimbra"/></soap:Header><soap:Body><soap:Fault><soap:Code><soa p:Value>soap:Receiver</soap:Value></soap:Code><soap:Reason><soap:Text>system failure: ZimbraLdapContext</soap:Text></soap:Reason><soapetail><Error xmlns="urn:zimbra"><Code>service.FAILURE</Code><Trace>com.zimbra.common.service.ServiceExcep tion: system failure: ZimbraLdapContext#012ExceptionId:btpool0-11://server.domain.com:7071/service/admin/soap/:1305553846588:cc5388fd07912327#012Code:service.FA ILURE#012#011at com.zimbra.common.service.ServiceException.FAILURE (ServiceException.java:248)#012#011at com.zimbra.cs.account.ldap.ZimbraLdapContext.&lt;i nit>(ZimbraLdapContext.java:423)#012#011at com.zimbra.cs.account.ldap.ZimbraLdapContext.&lt;i nit>(ZimbraLdapContext.java:374)#012#011at com.zimbra.cs.account.ldap.LdapProvisioning.getAcc ountByQuery(LdapProvisioning.java:641)#012#011at com.zimbra.cs.account.ldap.LdapProvisioning.getAcc ountByNameInternal(LdapProvisioning.java:809)#012# 011at com.zimbra.cs.account.ldap.LdapProvisioning.getAcc ountByName(LdapProvisioning.java:790)#012#011at com.zimbra.cs.account.ldap.LdapProvisioning.get(Ld apProvisioning.java:701)#012#011at com.zimbra.cs.account.ldap.LdapProvisioning.get(Ld apProvisioning.java:686)#012#011at com.zimbra.cs.service.account.Auth.handle(Auth.jav a:98)#012#011at com.zimbra.soap.SoapEngine.dispatchRequest(SoapEng ine.java:420)#012#011at com.zimbra.soap.SoapEngine.dispatch(SoapEngine.jav a:274)#012#011at com.zimbra.soap.SoapEngine.dispatch(SoapEngine.jav a:158)#012#011at com.zimbra.soap.SoapServlet.doWork(SoapServlet.jav a:291)#012#011at com.zimbra.soap.SoapServlet.doPost(SoapServlet.jav a:212)#012#011at javax.servlet.http.HttpServlet.service(HttpServlet .java:725)#012#011at com.zimbra.cs.servlet.ZimbraServlet.service(Zimbra Servlet.java:181)#012#011at javax.servlet.http.HttpServlet


As stated above a zmcontrol stop; zmcontrol start solves all problems


ulimit seems OK:

zimbra@dc1:~$ ulimit -n
524288


The customer is about to stop trial and revoke his order, so any help would be appreciated

[bdial]

if you're a network edition customer i would advise contacting support especially if it's this cdritical

[dvandenputte]

Indeed bdial.
Since it is a Trial Licence the support is not automatically active.
The Sales Rep is actually going to activate the support for this license.

BUT I believe I already find the solution

The /etc/security/limits.conf file already had:

zimbra soft nofile 524288
zimbra hard nofile 524288

I added :

root soft nofile 524288
root hard nofile 524288

and rebooted the server.

(from another post on the forum), and this probably solved the problem.

If confirmed by the customer I'll update this post (and quote the solving post)

[dvandenputte]

Indeed the problem was solved applying

Bug 42870 &ndash; "Bad file descriptor" SocketException causes system failure every several hours

Case closed.