我搜索了好久,但找不到太多关于这个问题的信息。pmie_daily 和 pmlogger_daily 服务都无法在我们的 RHEL 服务器上启动。它们是 RHEL“Performance Co-Pilot (PCP)”系统的一部分。我发现的一篇帖子说这是由于日志文件归 root 所有而不是 pcp 所有,但我检查了一下,我们的 pmie 日志文件归 pcp 所有。
服务日志只显示(pmie_daily 日志有类似的条目):
pmlogger_daily.service: Failed with result 'exit-code'.
pmlogger_daily.service: Main process exited, code=exited, status=1/FAILURE
Started Process archive logs.
而 /var/log/pcp/pmie/localhost/pmie.log 显示:
pmie: metric mssql.memory_manager.stolen_node_memory not in namespace for pmcd local: (host localhost)
pmLookupName failed: Unknown metric name
听起来很糟糕,但我不知道这意味着什么或如何解决它。
/var/log/pcp/pmlogger/pmlogger_daily.log 显示:
pmlogger_daily: [control.d/local:26]
Error: no pmlogger instance running for host "local:"
... logging for host "local:" unchanged
pmlogger_daily: [control.d/local:26]
Warning: skipping log rotation because we don't know which pmlogger to signal
pmlogger_daily: [control.d/local:26]
Warning: current volume of current pmlogger not known, compression skipped
/var/log/pcp/pmie/pmie_daily.log 显示:
pmie_daily: Error: no pmie instance running for host "local:"
[/etc/pcp/pmie/control.d/local:5] ... inference engine for host "local:" unchanged
...两者类似。这可能是问题所在吗?如果是,如何解决?