Ping failed on system when it can be pinged

Submitted by Exoware on Mon, 08/31/2009 - 05:38

I have an OpenVZ host system added, which I have thereafter added an OpenVZ instance from. Cloudmin reports that the ping has failed when in fact it is responding as I pinged it from the command line on the Cloudmin master. How does Cloudmin determine a host as not responding to ping?

Status: 
Closed (fixed)

Comments

Submitted by JamieCameron on Mon, 08/31/2009 - 12:16

Is the OpenVZ container perhaps paused? That can trigger this in the current Cloudmin release.

Submitted by Exoware on Mon, 08/31/2009 - 12:33

No, it was definitely not paused. It was running on the old system and I literally just added it into Cloudmin for management only (not moved it).

Submitted by JamieCameron on Mon, 08/31/2009 - 12:45

If you like, I can login to your Cloudmin system again and see what is going wrong here?

Submitted by Exoware on Mon, 08/31/2009 - 12:47

It's alright, once I moved it over, it was fine (and I don't have a system in the same situation anymore). I was just curious how Cloudmin determined a ping failure. Is there perhaps some way I can IM you for general discussion and for issues that don't really require an issue to be opened on this system?

Submitted by JamieCameron on Mon, 08/31/2009 - 12:50

For a ping failure, it generally just means that a command like :

ping -c 1 hostname

Failed for the OpenVZ container.

One other case in which this can fail is if the hostname doesn't resolve to the OpenVZ system's actual IP address.

By the way, you can IM me at jamiecameron@google.com

Submitted by Exoware on Mon, 08/31/2009 - 12:54

What IM protocol is that? Google Talk?

Submitted by JamieCameron on Mon, 08/31/2009 - 12:56

Yes, google talk..

Submitted by JamieCameron on Mon, 08/31/2009 - 13:33

For other readers - After discussion, this turned out to be caused by a mismatch between the hostname entered when the OpenVZ container was added, and its actual IP address.

Cloudmin will add better protection against this in the next release.

Submitted by Issues on Mon, 09/14/2009 - 17:19

Automatically closed -- issue fixed for 2 weeks with no activity.