[ZBX-1790] Oracle connection lost contact Created: 2010 Jan 22  Updated: 2017 May 30  Resolved: 2010 Feb 10

Status: Closed
Project: ZABBIX BUGS AND ISSUES
Component/s: Server (S)
Affects Version/s: 1.8.2, 1.9.0 (alpha)
Fix Version/s: 1.8.2, 1.9.0 (alpha)

Type: Incident report Priority: Blocker
Reporter: Alexey Pustovalov Assignee: Alexander Vladishev
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Oracle 10g2, Gentoo 2.6.31-gentoo-r6

Issue Links:
Duplicate
duplicates ZBX-1945 Speedup synchronization between nodes Closed

 Description   

In log:
11403:20100121:223427.071 server #175 started [Node watcher. Node ID:0]
11403:20100122:184616.881 [Z3001] Connection to database '(null)' failed: [-1] ORA-03135: connection lost contact
11140:20100122:184617.023 One child process died (PID:11403). Exiting ...

Parameter Value Details
Zabbix server is running Yes -
Number of hosts (monitored/not monitored/templates) 611 472 / 12 / 127
Number of items (monitored/disabled/not supported) 25544 24600 / 944 / 0
Number of triggers (enabled/disabled)[true/unknown/false] 9720 9641 / 79 [96 / 4226 / 5319]
Number of users (online) 48 4
Required server performance, new values per second 69.071549272487 -

thise error every week and any time, maybe few server crash on week.

 Comments   
Comment by Alexei Vladishev [ 2010 Jan 29 ]

Probably we should handle it as we do it for MySQL.

Comment by Jan Hrebenar [ 2010 Feb 09 ]

Hi,

I'm having the same problem. Few times a week we have this error. I suspecting the full RMAN backup that cause a big load on DB and IO.

I found also an warning in DB alert_log at the time of zabbix crash "WARNING: inbound connection timed out (ORA-3136)"

Some discussion about the problem here: http://forums.oracle.com/forums/thread.jspa?messageID=1612957&tstart=0

Regards,

oldsam

Comment by Alexei Vladishev [ 2010 Feb 10 ]

I am closing this issue. See also ZBX-1945, these issues are related to each other.

It was fixed so that once started, Zabbix Node Syncer process will not reconnect to the database any more. In 1.8 and 1.8.1 Node Syncer process was running 1) regardless of distributed setup and 2) periodically connected and disconnected from the database.

Now it no longer runs in non DM setup, also it connects to the database on startup and never disconnects.

Alexei

Generated at Sat Dec 13 21:48:20 EET 2025 using Jira 10.3.13#10030013-sha1:56dd970ae30ebfeda3a697d25be1f6388b68a422.