On Tue, 31 Oct 2000, Alan Bailey wrote:
>
> Did the format of the pmlogger control file get changed without the
> documentation getting changed?
Absolutely not, unless someone has a strong death wish!
> I'm setting up a logging machine to log a few metrics from a bunch of
> hosts. When I have the control file setup the way the documentation
> explains, it grabs one of the 'n's as the directory name, and creates it.
> It then uses 'n' for the directories for each host, which is invalid also.
> Should there just be one 'n' in the config file? It seems to work that
> way.
Could you please send me your /var/pcp/config/pmlogger/control.cluster
file as an attachment so I can check for whitespace strangeness?
> I'm running pcp-2.1.10 on RH 6.2.
That is a known good combination.
Some other things that would help ...
1. output from
$ for f in /var/tmp/pmlogger/*; do echo === $f ===; cat $f; done
before and after running /usr/share/pcp/bin/pmlogger_check
2. output from
$ /usr/share/pcp/bin/pmlogger_check -NV
3. output from
$ sh -x /usr/share/pcp/bin/pmlogger_check
> Thanks for any help,
> Alan
>
> output below:
> ------------
>
> [root@lanner abailey]# cat /var/pcp/config/pmlogger/control.cluster
> ntsc1033 n n PCP_LOG_DIR/pmlogger/ntsc1033 -c config.cluster
> ntsc1034 n n PCP_LOG_DIR/pmlogger/ntsc1034 -c config.cluster
> ntsc1035 n n PCP_LOG_DIR/pmlogger/ntsc1035 -c config.cluster
> [root@lanner abailey]# /usr/share/pcp/bin/pmlogger_check -c
> /var/pcp/config/pmlogger/control.cluster
> pmlogger_check [/var/pcp/config/pmlogger/control.cluster:1]
> Warning: creating directory (n) for PCP archive files
OK, this looks wrong, but your control file looks OK.
> Restarting pmlogger for host "ntsc1033" ... timed out waiting!
This looks like a second problem ...
> Usage: pmlogger [options] archive
>
> Options:
> -c configfile file to load configuration from
> -h host metrics source is PMCD on host
> -l logfile redirect diagnostics and trace output
> -L linger, even if not primary logger instance and
> nothing to log
> -n pmnsfile use an alternative PMNS
> -P execute as primary logger instance
> -r report record sizes and archive growth rate
> -s endsize terminate after endsize has been accumulated
> -t interval default logging interval [default 60.0 seconds]
> -T endtime terminate at given time
> -v volsize switch log volumes after volsize has been
> accumulated
> -V version generate version 1 or 2 archives (default is 2)
> -x fd control file descriptor for application launching
> pmlogger
> via pmRecordControl(3)
> pmlogger_check: Error: cannot find pmlogger output file at "pmlogger.log"
> Directory (/home/abailey/n) contents:
> total 8
> drwxrwxr-x 2 root root 4096 Oct 31 14:09 .
> drwx------ 18 abailey abailey 4096 Oct 31 14:09 ..
> ---------- 1 root root 0 Oct 31 14:09 lock
> pmlogger_check: Error: archive file 20001031.14.09.0 missing
> Directory (/home/abailey/n) contents:
> total 8
> drwxrwxr-x 2 root root 4096 Oct 31 14:09 .
> drwx------ 18 abailey abailey 4096 Oct 31 14:09 ..
> ---------- 1 root root 0 Oct 31 14:09 lock
> pmlogger_check [/var/pcp/config/pmlogger/control.cluster:2]
> Warning: creating directory (n) for PCP archive files
> pmlogger_check: [/var/pcp/config/pmlogger/control.cluster:2]
> Error: Cannot start more than one pmlogger instance for archive directory
> "n"
> ... logging for host "ntsc1034" unchanged
> pmlogger_check: [/var/pcp/config/pmlogger/control.cluster:3]
> Error: Cannot start more than one pmlogger instance for archive directory
> "n"
> ... logging for host "ntsc1035" unchanged
> [root@lanner abailey]#
>
>
|