On Thursday, June 30, 2016 at 8:26:50 PM UTC+3, Michael G. wrote: > Recently I've been revisiting influxdb and trying to setup some toy > deployments. As part of this process we've set up a handful of systems with > collectd pulling metrics and pushing to my instance of influxdb over the > collectd network plugin. There are several examples of this setup available > but it seems I can't get data from the remote instances of collectd to > populate in influx. The catch here is that a local collectd instance that > monitors the influxdb system is able to populate but it's the only one. Below > are the debug steps I've taken, any advice or something I might be > overlooking is appreciated. > > > system information: > CentOS Linux release 7.2.1511 > collectd-5.5.0 > influxdb-0.13.0.x86_64 > > > influxdb configuration: > > > > [[collectd]] > enabled = true > database = "collectd" > typesdb = "/usr/share/collectd/types.db" > bind-address = ":25826" > collectd network plugin configuration: (same on all remote/local) > > > > <Plugin network> > Server "metrics.tst" "25826" > </Plugin> > > I've verified the following for debug purposes as per various guides online: > > > UDP buffer size: > > > > [root@metrics ~]# sysctl net.core.rmem_max > net.core.rmem_max = 8388608 > Verify receiving packets: > > > > [root@metrics ~]# tcpdump -v -n dst port 25826 > tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 65535 > bytes > 13:06:47.283486 IP (tos 0x0, ttl 63, id 0, offset 0, flags [DF], proto UDP > (17), length 1339) > 10.0.0.100.46339 > 10.0.0.10.25826: UDP, length 1311 > 13:06:49.283651 IP (tos 0x0, ttl 63, id 0, offset 0, flags [DF], proto UDP > (17), length 1336) > 10.0.0.100.46339 > 10.0.0.10.25826: UDP, length 1308 > Verify that I'm listening: > > > > > [root@metrics ~]# ss -anu > State Recv-Q Send-Q Local Address:Port Peer Address:Port > UNCONN 0 0 :::25826 :::* > > > > [root@metrics ~]# netstat -anlp | grep 25826 > udp6 0 0 :::25826 :::* > 9387/influxd > > influxdb log output: > > > > [root@metrics ~]# tail -F /var/log/influxdb/influxd.log | grep collectd > [collectd] 2016/06/30 13:15:01 Starting collectd service > [collectd] 2016/06/30 13:15:01 Listening on UDP: [::]:25826 > > Every time when I try to dump the series information: > > > > [root@metrics ~]# influx > Visit https://enterprise.influxdata.com to register for updates, InfluxDB > server management, and monitoring. > Connected to http://localhost:8086 version 0.13.0 > InfluxDB shell version: 0.13.0 > > use collectd > Using database collectd > > show series > > > > However when I use a local collectd instance I get several populated > series.I've tried looking for more debugging output from influx but what the > collectd service provides isn't much. Additionally the I haven't found > anything in stats or diagnostics that helps. Any ideas anyone? > Regards, > -Michael
Hi Michael, I suggest opening up a dummy port on your "metrics" server to see how the incoming udp traffic looks like. An easy way to do this is using netcat. After stopping influxdb, start listening on the UDP port 25826 with nc and examine your UDP input: `nc -u -l 25826` This would be exactly what the InfluxDB collectd plugin receives. Does the input look right? Thanks, Kostas -- Remember to include the InfluxDB version number with all issue reports --- You received this message because you are subscribed to the Google Groups "InfluxDB" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/influxdb. To view this discussion on the web visit https://groups.google.com/d/msgid/influxdb/84ded8d7-1f96-4a25-a047-3b41f2e5bf8b%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
