[Xymon] Very peculiar behaviour in Analysis.cfg

[Neil Simmonds]

Hi all,
I seem to be having problems with CLASS in analysis.cfg and I'm really not
sure as to why. I hope I can explain this well enough.

I have some servers picking up the rules from analysis.cfg and some not,
I've selected 2 as samples to try and illustrate the issue.

These are the entries from hosts.cfg for each server

172.01.01.01    server1 # http://server1:2801/utils/enterLogin.jsp
NOPROPYELLOW:disk,cpu,mq delayyellow=cpu:15 delayred=cpu:15 CLASS:linux
172.01.01.02    server2 # NOPROPYELLOW:disk,cpu delayyellow=cpu:15
delayred=cpu:15 CLASS:linux

As you can see, both force the CLASS:linux parameter.

Each server has it's own config file in an analysis.d directory which is
included near the bottom of the analysis.cfg file.

server 1's file contains this,

HOST=server1 GROUP=unixsupport
        PORT "LOCAL=%([.:]22)$" state=LISTEN TEXT=ssh
        PROC cybAgent.bin 1
        PROC rmi_server1 1 -1 red
        PROC admin_server1 1 -1 red
        PROC eventprocessor_server1 1 -1 red
        PROC queuemanager_server1 1 -1 red
        PROC scheduler_server1 1 -1 red
        PROC workflowengine_server1 1 -1 red
        PROC "server1Cell01 server1Node01 pim01srv" 1 -1 red
        PROC init 1 -1 red
        PROC tlmagent.bin 1
        PROC /opt/tivoli/tsm/client/ba/bin/dsmcad 1
        PROC "/opt/tivoli/tsm/client/ba/bin/dsmc sched" 1
        LOG %/opt/IBM/MDMPIM/logs/\w*server1/\w*.log "internal error"
COLOR=yellow IGNORE="unregistered"
        LOG /opt/IBM/MDMPIM/logs/workflowengine_server1/db_manager.log
"CWPCM0033W:An error occurred when registering thread:
heartbeat_workflowengine" COLOR=red
        LOG /applogs/mdmpim/logs/appsvr_server1/default.log
"/utils/custom_page.jsp] ERROR jsp.footer - CWPCM0002E:Generic error /
Exception: Not an AustinException, Exception:Failure to acquire exclusive
lock for AustinContext" COLOR=red
        MQ_QUEUE SYSTEM.DEAD.LETTER.QUEUE depth-warning=1
        MQ_QUEUE "%([A-Z.]*BO)$" depth-warning=1

and server 2's contains this,

HOST=server2 GROUP=unixsupport
        PORT "LOCAL=%([.:]22)$" state=LISTEN TEXT=ssh
        PROC /home/dasusr1/das/adm/db2dasrrm 1
        PROC /opt/IBM/db2/V9.7/bin/db2fmcd 1
        PROC /opt/IBM/db2/V9.7/das/bin/db2fmd 1
        PROC db2sysc
        PROC tlmagent.bin 1
        PROC /opt/tivoli/tsm/client/ba/bin/dsmcad 1
        PROC "/opt/tivoli/tsm/client/ba/bin/dsmc sched" 1

Nothing unusual or out of the ordinary in either of those.

in my analysis.cfg file I have the following,

HOST=*
        DISK %^/mnt IGNORE
        DISK %^/nimdir 101 101
        DISK %^/media IGNORE
        DISK %^/cdrom IGNORE

CLASS=win32
#       SVC CentennialIPTransferAgent startup=automatic status=started
#       SVC %McShield startup=automatic status=started
#       CLOCK 600 yellow

CLASS=linux
        UP      10m
        PROC ntpd 1
        PROC /usr/sbin/sshd 1
        DISK /nmon 95 98 TIME=W:0800:1800
        DISK /nmon 95 101


CLASS=aix
#       PROC /usr/sbin/syslogd 1
        PROC /usr/sbin/xntpd 1
        PROC /usr/sbin/cron 1
        PROC /usr/lib/errdemon 1
        PROC /usr/sbin/syncd 1
        PROC sshd 1
        DISK /nmon 95 98 TIME=W:0800:1800
        DISK /nmon 95 101 TIME=W:1800:0800
        DISK /perf 95 98 TIME=W:0800:1800
        DISK /perf 95 101 TIME=W:1800:0800
        DISK /temp IGNORE

# include files
directory /xymon/server/etc/analysis.d

CLASS=linux
        MEMSWAP 10 30

DEFAULT
     # These are the built-in defaults.
     UP      10m
     LOAD    6.0 8.0
     DISK    * 95 98
     MEMPHYS 100 101
     MEMSWAP 50 80
     MEMACT  90 97


Server 1 does not pick up the lines in the analysis.cfg file yet server 2
does. That is the confusing thing, as far as I can see, both should pick up
these rules.

Thanks,
Neil.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.xymon.com/pipermail/xymon/attachments/20151012/12061d97/attachment.html>