If MWA is not working, MeasureWare processes "scopeux", "perflbd" and "midaemon"
1. run "mwa start", after that check whether all monitored
MeasureWare processes are running:
# mwa start
# ps -e | grep -e scopeux -e perflbd -e midaemon
You should see the output like this :
20948 ? 0:00 perflbd
20940 ? 0:00 scopeux
20937 ? 0:00 midaemon
If you can see "perflbd","scopeux" and "midaemon" processes in the output, stop
here.
If you can not see any of the 3 processes, go to the following steps.
2. Stop all MeasureWare process
1). stop measure process
# mwa stop
2). stop ttd process
# ttd -k
3). Sometimes the above 2 steps can not stop all measureware processes.
Verify this with command "ps -ef grep /opt/perf", if any measureware
processes still running, kill them.
For example :
# ps -ef |grep perf
root 1657 1 0 Apr 15 ? 124:28 /opt/perf/bin/perflbd
#
# kill 1657
If the above "kill" command can not kill the process, please try it with
option "- 9"
# kill -9 1657
3. backup all performance data
For example :
# mkdir /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logappl /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logdev /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logglob /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logindex /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logproc /var/adm/perf/datafile/old_logs
or you can backup the above files to some other directory you want to.
If the log files are too huge, please also gzip the backup files.
4. Zero-size all performance data
> /var/opt/perf/datafiles/logappl
> /var/opt/perf/datafiles/logdev
> /var/opt/perf/datafiles/logglob
> /var/opt/perf/datafiles/logindex
> /var/opt/perf/datafiles/logproc
5. Start MWA again
# mwa start
6. Ensure all monitored process are running :
# ps -eaf |grep -e scopeux -e perflbd -e midaemon ; date
You should see the output like this :
20948 ? 0:00 perflbd
20940 ? 0:00 scopeux
20937 ? 0:00 midaemon
If you can see "perflbd","scopeux" and "midaemon" process in the output,
you are good here
1. run "mwa start", after that check whether all monitored
MeasureWare processes are running:
# mwa start
# ps -e | grep -e scopeux -e perflbd -e midaemon
You should see the output like this :
20948 ? 0:00 perflbd
20940 ? 0:00 scopeux
20937 ? 0:00 midaemon
If you can see "perflbd","scopeux" and "midaemon" processes in the output, stop
here.
If you can not see any of the 3 processes, go to the following steps.
2. Stop all MeasureWare process
1). stop measure process
# mwa stop
2). stop ttd process
# ttd -k
3). Sometimes the above 2 steps can not stop all measureware processes.
Verify this with command "ps -ef grep /opt/perf", if any measureware
processes still running, kill them.
For example :
# ps -ef |grep perf
root 1657 1 0 Apr 15 ? 124:28 /opt/perf/bin/perflbd
#
# kill 1657
If the above "kill" command can not kill the process, please try it with
option "- 9"
# kill -9 1657
3. backup all performance data
For example :
# mkdir /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logappl /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logdev /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logglob /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logindex /var/adm/perf/datafile/old_logs
cp -p /var/opt/perf/datafiles/logproc /var/adm/perf/datafile/old_logs
or you can backup the above files to some other directory you want to.
If the log files are too huge, please also gzip the backup files.
4. Zero-size all performance data
> /var/opt/perf/datafiles/logappl
> /var/opt/perf/datafiles/logdev
> /var/opt/perf/datafiles/logglob
> /var/opt/perf/datafiles/logindex
> /var/opt/perf/datafiles/logproc
5. Start MWA again
# mwa start
6. Ensure all monitored process are running :
# ps -eaf |grep -e scopeux -e perflbd -e midaemon ; date
You should see the output like this :
20948 ? 0:00 perflbd
20940 ? 0:00 scopeux
20937 ? 0:00 midaemon
If you can see "perflbd","scopeux" and "midaemon" process in the output,
you are good here
No comments:
Post a Comment