topic Nimble Heartbeats dropping in Array Setup and Networking

Nimble Heartbeats dropping

xpystchrisx82 — Mon, 20 Oct 2014 12:08:14 GMT

We've had our CS260G in play for a number of months now and everything is awesome. Love the performance, upgrades, pretty much everything has been exactly as we had expected.

We are encountering one issue which has me scratching my head. Approximately once every week InfoSight sends a "Nimble Storage no longer recieving array heartbeats" in reference to our CS260G. To date there has never been a single problem, and manually sending array autosupport connections always sends an automated ticket close. Has anyone else experienced this? We're scratching our heads. All of our network gear shows the traffic to the InfoSight IP address (198.54.168.99) as having been allowed. I've never had an issue downloading Nimble software or getting to InfoSight, Connect, or the Corporate website.

Re: Nimble Heartbeats dropping

rugby0134 — Mon, 20 Oct 2014 14:13:56 GMT

I've seen this before and it ending being a network cable that was not seated. I would re-seat the cables of the standby controller, or replace them. Then fail-over and repeat.

Re: Nimble Heartbeats dropping

xpystchrisx82 — Mon, 20 Oct 2014 14:24:58 GMT

Any chance we can perform the cable replacement without a fail over? I have two cables plugged into the management ports of each node.

Re: Nimble Heartbeats dropping

rugby0134 — Mon, 20 Oct 2014 14:36:31 GMT

Yes - you can - but if you take too long - it's going to fail-over the ctrl as a network down. The alert should have info on which controller is having the issue. You can just fix the cables on that controller.

Re: Nimble Heartbeats dropping

xpystchrisx82 — Tue, 21 Oct 2014 20:27:46 GMT

I believe we have some network congestion, or one of our security appliances is intermittently blocking the InfoSight IP address. Yesterday I wasn't able to connect to InfoSight and on-queue I got an email saying that the system heartbeats were dropped. Digging through our network to try and figure out where the issue actually lies. I'll post up if/when I figure out the problem.