2010/11/29 Sofer, Yuval <Yuval_Sofer(at)bmc(dot)com>:
> He reported about storage problems which triggered node switch (it is postgres installation on windows 2008 cluster)
> But now although postgres.exe is running, it is not really functioning…
>
Sounds to me like your clustering software caused some data
loss/corruption. Have you ever tested cluster failover before putting
it in production? Did it cause this problem then?