Problem with oracle conection

[efren.gl]

hi,

Currently I have my Zabbix 1.6.5 with oracle 10g database, Zabbix suddenly appears as "ZABBIX server is running NO", so search the log and I see him following error:

16645:20090718:080146 One child process died. Exiting ...
16645:20090718:080148 [Z3001] Connection to database 'zabbix/password@SID_ORACLE' failed: [0] sqlo_connect

Can anyone tell me that will help and this happens with Oracle?

Best regards, Efren

[untergeek]

Oracle connection problem

2 things we've hit with our config:

1. You may need more listeners. More listeners means more oracle connections. You can fix this by basically increasing the numbers in your zabbix_server.conf file

2. Your database ran out of space, even temporarily, in an index file, in a redo log space, in something.

We fixed a cron based cleanup script to ensure that we didn't run out of space, and updated the number of listeners/trappers/pingers, etc. We haven't seen this problem since.

UPDATE: CORRECTION: RETRACTION. As I was writing this, our zabbix_server died from this exact error. We may still have had the situation where we ran out of redo log space (this did happen). I don't know yet.

[efren.gl]

Oracle connection problem

Thank you untergeek

What is the name of the parameter to (listeners) zabbix_server.conf defined file, what happens
is what I tried but no finds.

Best regards.

[untergeek]

If you read my correction at the bottom, it appears my "solution" isn't really solving anything.

That said, here's a list of attributes which you can increase. Be careful not to over-do it. You can make the system just as unstable with high numbers as with low ones.

# Number of pre-forked instances of pollers
# Default value is 5
# This parameter must be between 0 and 255
StartPollers=12

# Number of pre-forked instances of IPMI pollers
# Default value is 0
# This parameter must be between 0 and 255
StartIPMIPollers=0

# Number of pre-forked instances of pollers for unreachable hosts
# Default value is 1
# This parameter must be between 0 and 255
StartPollersUnreachable=3

# Number of pre-forked instances of trappers
# Default value is 5
# This parameter must be between 0 and 255
StartTrappers=12

# Number of pre-forked instances of ICMP pingers
# Default value is 1
# This parameter must be between 0 and 255
StartPingers=3

# Number of pre-forked instances of discoverers
# Default value is 1
# This parameter must be between 0 and 255
StartDiscoverers=1

# Number of pre-forked instances of HTTP pollers
# Default value is 1
# This parameter must be between 0 and 255
StartHTTPPollers=3

Note that the numbers are higher than the defaults.

Good luck!

[sj7trunks]

Oracle dies

I've noticed this problem in my environment too. Server can stay up for 10-15 minutes before crashing, in the source code it looks as if the next function would be to wait 10 seconds until the database comes back. This behavior is not noticed and directly crashes before waiting.

Filed a bug: https://support.zabbix.com/browse/ZBX-1123