On Wed, Aug 11, 2010 at 3:33 PM, David Lang <[email protected]> wrote: > On Wed, 11 Aug 2010, Igor Chudov wrote: > >> A closer look at the ha-debug files, that I posted here an hour ago, >> shows that there is a bit of extra detail about reading config files >> and then all the same stuff as in ha-log. >> >> At this point, I am beginning to have my doubts about this whole >> heartbeat system and its ability to serve for years, in what looks to >> me like simple configuration. >> >> Even more than the fact that it does not work, I am bugged that it is >> essentially not transparent, not troubleshootable, and if something >> goes wrong with it, it will be hard to troubleshoot under stress. >> >> I mean, I am really asking a basic question, why exactly do they >> decide to become, or not become, masters? And there is no answer to >> this most important questions in the logs. Just nothing. >> >> The "corosync" project is also unmanageable, having a GUI that does >> not work, completely opaque XML based configuration, and extremely >> poor documentation. >> >> The reason for trying to create this setup is that we have a three >> year old setup, also based on drbd and heartbeat. This system is >> having "trouble" because somehow or other, it became a lot slower than >> it has been. That is so, even though the hardware is just fine. >> Everything just became slower. >> >> In a somewhat similar vein to the current situation, this old system >> is quite resistant to troubleshooting as well. >> >> I think that I will do a complete OS reinstall on these two new >> machines and will try again, but at this point I am much less >> optimistic. >> >> I would like to say that I am very thankful for your help and still >> have some hopes, but much less optimistic than I was. >> >> So, are there any simple alternatives to what I am trying to do? > > Igor, > I have been using heartbeat for about a decade now and it has been very > reliable. > > the problem here is finding where the logs are going. > > you had one batch of logs that showed exactly the type of thing that I > expected > to see (the ones where you started heartbeat 10 min later and then something > external to heartbeat shut it down) > > this is why I asked you to check the logd.cf to see if it's sending logs > elsewhere. For example, it could be sending them to syslog and if we look > there > we will find what we need. >
Here's my logd.cf, same on both machines: r...@pfs-srv3:~# cat /etc/logd.cf debugfile /var/log/ha-debug logfile /var/log/ha-log r...@pfs-srv4:~# cat /etc/logd.cf debugfile /var/log/ha-debug logfile /var/log/ha-log > the other option is to disable logd and define logfile (if it's still > supported > in the 3.x version you are running). i.e. something like 'logfile > /var/log/ha-log' > > > > > if we can't see the logs of what's happening, it is impossible to > troubleshoot. > You did have one set of files that showed what is happening, what was > different > about that run. Well, I can repost ha-debug and ha-log files. I have them and I posted them. Just let me know if you would like to see them again and I will resend them. Igor > I sympathise with the problems you are having with your old system, but it > becoming slower is almost certinly not related to heartbeat. > > David Lang > > _______________________________________________ > Linux-HA mailing list > [email protected] > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
