4.2 hobbitd_alert EXPAGE= bug?

John Glowacki johng at idttechnology.com
Mon Apr 2 21:15:17 CEST 2007


Is there a bug with EXPAGE=? When I run a --test it excludes the page,
but hobbit runs the alert script. page.log seems to confirm it. I am
running 4.2 with allinone patch. Humm. More details as I look into this.
I have 4 testdb systems in alert on the page. testdb02/3/4/7. Only
testdb03.disk and testdb04.adrian is being excluded. Any idea why?

# ../bin/bbcmd hobbitd_alert --dump-config
2007-04-02 11:37:26 Using default environment file
/opt/hobbit/server/etc/hobbitserver.cfg
EXPAGE=%test EXHOST=newman
        SCRIPT /opt/hobbit/server/etc/alert.sh noc FORMAT=SCRIPT REPEAT=3

# ../bin/bbcmd hobbitd_alert --config=hobbit-alerts.cfg --test  testdb02
 disk
2007-04-02 11:49:49 Using default environment file
/opt/hobbit/server/etc/hobbitserver.cfg
00011109 2007-04-02 11:49:49 send_alert testdb02:disk state Paging
00011109 2007-04-02 11:49:49 Matching host:service:page
'testdb02:disk:test' against rule line 120
00011109 2007-04-02 11:49:49 Failed 'EXPAGE=%test EXHOST=newman
RECOVERED NOTICE SCRIPT /opt/hobbit/server/etc/alert.sh noc
FORMAT=SCRIPT REPEAT=3' (pagename excluded)

alert.sh script output shows 3 of the 4 systems alerting:
# cat  /var/log/hobbit_alerts.log  | grep testdb0 | awk '{ printf "%s %s
%s %s %s\n", $6, $7, $8, $9, $10 }' | sort | uniq -c
     60 red testdb02 disk 0 120
     60 red testdb04 vxvm 0 120
     60 red testdb07 vxvm 0 120

# grep testdb0 notifications.log | awk '{ print $6 }' | sort -u
testdb02.disk
testdb04.vxvm
testdb07.vxvm

page.log
2007-04-02 12:16:20 send_alert testdb02:disk state 0
2007-04-02 12:16:20 Found a first matching rule
2007-04-02 12:16:20   repeat testdb02|disk|script|noc at 1175530552
2007-04-02 12:16:20   Alert for testdb02:disk to noc

2007-04-02 14:14:11 Found no first matching rule
2007-04-02 14:14:11 cleanup_alert called for host testdb04, test adrian
2007-04-02 14:17:15 Found no first matching rule
2007-04-02 14:17:15 cleanup_alert called for host testdb03, test disk

Thanks,
John



More information about the Xymon mailing list