Hi, I am facing a weird issue here. we have a Solaris 10 server running 15 non-global zones. The issue is all the systems suddenly stopped connectivity from outside/remotely..
Ping is also not working.
After some troubleshooting, what I observed is when all the non-global zones are halted, the global -IP is reachable fine with no issues. As soon as all the non-global zones started, we are facing intermittent network disconnect for some time then it will stop the connectivity completely after some time..
Please let me know what can be the fix and how can I stop this..
Wow, that does sound weird.
My initial questions are:
- Are all the non-global zones Solaris 10? Or are there branded zones too?
- Are the non-global zones whole root or sparse root?
- Are they sharing network interfaces? Or are some exclusive network?
You say that, running on its own, the global zone is normal. So I would suggest that you prevent the non-globals from starting automatically, and start them one at a time and find when the trouble starts.
I'd be inclined to examine (even print off) each zone configuration and compare them. Ensure all ip addresses are unique.
A pure guess would be that something may be upsetting an intelligent network switch that, as a result, throws its toys out of the pram and decides not to play any more. That could provide the impression of total network failure of all zones but not be caused (directly) by the Solaris machine itself but the mis-configuration of a zone..
I think a step-by-step logical approach is needed, starting zones one at a time.
1 Like
Never mind. we found out that there was an incoming ping storm.. We blocked that IP and we are good now..
Thanks for looking into it.
Just out of curiosity, what did you look for to discover this, Snchaudhari2?
Checked for snoop data on each interface..
1 Like