Hi Pete, this error can also be triggered by a large clock skew between the systems. Are you using NTP (or any similar technology) to sync the clocks of the systems running Graylog?
Cheers, Jochen On Thursday, 6 August 2015 23:33:33 UTC+2, Pete GS wrote: > > One of my nodes this morning is reporting the meta info error in the logs > so I took the opportunity to look through the API browser again. > > Interestingly enough all the MongoDB stats look fine and normal in there, > no HTTP 500 status or anything like that. > > Aside from monitoring the Graylog server log for these events I'm stumped > what else to look for at present... > > Cheers, Pete > > On Fri, Jul 31, 2015 at 6:55 PM, Jochen Schalanda wrote: > >> Hi Pete, >> >> the MongoDB stats resource will respond with HTTP status 500 after the >> timeout for connecting the MongoDB server has been reached (default: 10s) >> if the MongoDB server is down. >> >> Cheers, >> Jochen >> >> On Thursday, 30 July 2015 23:46:51 UTC+2, Pete GS wrote: >>> >>> Thanks Jochen, I thought I'd navigated through pretty much all the API >>> Browser yesterday but I completely missed those! >>> >>> I'll check the Mongo stats next time it happens and see what I can see >>> in there compared to a connected node and that should hopefully do the >>> trick. >>> >>> Cheers, Pete >>> >>> On Thursday, 30 July 2015 19:45:27 UTC+10, Jochen Schalanda wrote: >>>> >>>> Hi Pete, >>>> >>>> currently there is no resource in the Graylog REST API which would >>>> check the availability of MongoDB or Elasticsearch explicitly. But you >>>> could check this "indirectly" via the cluster stats resource at >>>> http://localhost:12900/system/cluster/stats (or more specifically >>>> http://localhost:12900/system/cluster/stats/elasticsearch and >>>> http://localhost:12900/system/cluster/stats/mongo). >>>> >>>> Cheers, >>>> Jochen >>>> >>>> On Thursday, 30 July 2015 02:12:19 UTC+2, Pete GS wrote: >>>>> >>>>> This is possibly a little obscure but also possibly useful... >>>>> >>>>> I've written a Nagios plugin (in Perl) to check the health of all my >>>>> Graylog nodes but the one thing I can't seem to find how to check is the >>>>> status of a Graylog node in relation to being able to connect to the >>>>> MongoDB. I can check pretty much everything else I want to (eg. journal >>>>> utilisation, messages in vs. out, etc.). >>>>> >>>>> Essentially we have some funky network issues on occasion that will >>>>> stop one or more nodes from talking to the MongoDB servers and they don't >>>>> always recover, meaning one or more nodes will constantly report "Did not >>>>> find meta info of this node. Re-registering." in the server.log. >>>>> >>>>> I can certainly run another stream in Graylog and alert on this but >>>>> I'd much prefer to be able to get it from the API if this is possible. >>>>> >>>>> Is there a function in the API already for this that I'm missing or >>>>> should this be a feature request? >>>>> >>>>> I realise the key here is fixing the funky network issues and I'm >>>>> working on that in parallel :) >>>>> >>>>> Cheers, Pete >>>>> >>>> > -- You received this message because you are subscribed to the Google Groups "Graylog Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/graylog2/6bcc6ecc-9746-4c53-acb3-c89bdfc73f6e%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
