Mark Goodwin

stopped all pmloggers and pmproxy. Restored the pre-pmlogger_daily pmlogger directory, then restarted all pmloggers. Once they were all started, manually ran pmproxy under valgrind. Then removed pmlogger_daily.stamp and manually started...

Clearing assignment - cant repro without large pmlogger farm deployment

If pmDiscoverInvokeValuesCallBack was responsible for pmFreeResult (instead of process_logvol), then running the code async should work fine - you can throttle the redis requests however you like without having to...

In particular, most calls to redisSlotsRequest do not check the return value (especially for -ENOMEM). Related to https://github.com/performancecopilot/pcp/pull/1301

Nathan mentioned the fix probably requires an additional pass over the parse tree.

Note qa/1886 has some subtests that will need to be remade once this issue is fixed. e.g. Verify min/max() functions for a non-sigular metric, only one instance `pmseries $args 'kernel.all.load{instance.name...

Notifying systemd as soon as the primary has started was intentional - we can't (and don't need to) wait for the whole farm to start during boot. The TODO for...

This is a bug and we should just fix it. It's not only the PMLOGGER_INTERVAL default value that's missing from the environment of non-primary loggers. re: adding a new PMLOGGER_REMOTE_INTERVAL...

re: "changing behavior which has been there for 6 years" => viewed as a regression to fix the bug. Any further thoughts anyone?

In addition to `PMLOGGER_INTERVAL` not being set in the environment of secondary loggers, the following are also missing: `PMLOGGER_LOCAL`, `PMLOGGER_MAXPENDING`, `PMLOGGER_CHECK_SKIP_LOGCONF` `PMLOGGER_CHECK_SKIP_LOGCONF` not being set leaves loggers that are expecting...