Re: [Bug 1359975] pmchart run-away mem leak replaying multi-archive when rewinding
"Nathan Scott" <[email protected]>
| Newsgroups | gmane.comp.sysutils.pcp |
|---|---|
| Message-ID | <[email protected]> |
----- Original Message ----- > On 16/12/16 11:28, Nathan Scott wrote: > > ... > > ... at the point that Dave was hacking in libpcp, all instances seemed > > to have been sorted at some point along the pipeline before that point. > > But we couldn't identify where that was (if it was indeed the case) - so > > the above metric (logged, and then observed at the point of interest) is > > hopefully going to provide an acid test for that case. > > Metrics delivered by pmFetchGroup() will have sorted instances, I think, > because pmFetchGroup() calls pmSortInstances(). (fetchgroup is not in play here, AFAIK; optfetch was possibly though, via pmlogger) > Metrics delivered by pmFetch() won't get sorted (other than the order > they're delivered by pmcd, which does not appear to call qsort anywhere). Yep. There's also a case where the archive was created by libpcp_import which I think may actually be the case here (Mark?) - as it also makes a call to pmSortInstances. However, in libpcp where Dave is working we can't assume sorting I think, unless pmlogger ensures sorting before writing (and I don't think it does). cheers. -- Nathan -=-=-=-=-=-=-=-=-=-=-=- Groups.io Links: You receive all messages sent to this group. View/Reply Online (#14916): https://groups.io/g/pcp/message/14916 View All Messages In Topic (8): https://groups.io/g/pcp/topic/3803079 Mute This Topic: https://groups.io/mt/3803079?uid=174580 New Topic: https://groups.io/g/pcp/post Change Your Subscription: https://groups.io/g/pcp/editsub?uid=174580 Group Home: https://groups.io/g/pcp Contact Group Owner: [email protected] Terms of Service: https://groups.io/static/tos Unsubscribe: https://groups.io/g/pcp/leave/354243/563757577/xyzzy -=-=- pcp mailing list [email protected] https://groups.io/g/pcp/messages -=-=-=-=-=-=-=-=-=-=-=-