Well, I think that the issues may be multiple- For starters, yesterday they were doing some tests on the core network and all 4 hosts lost their network connectivity- HA initiated, and each ended up going into isolation mode and powering down all of their guests. When network was restored, I verified that each host could ping their gateway but after powering up one of the guest, found that it (the guest) could not ping its gateway. After powering up another guest, same symptoms however the guests could ping one another. A call to VMware essentially shifted focus on the physical switch (p-class enclosure) as well. After 2.5 hours of, things starting to auto-magically work once the speed on a hosts physical nic was changed from 1000Mb down to 10Mb.
A second issue is when a guest gets vmotion'd to another host they randomly lose their network connectivity. Only after disconnecting/re-connecting the network adapter in the guests vm settings, does the guest network restore. And a smaller percentage of the times even this fails - but a migration of the guest to another host fixes these few that the previous fix fails on.
Not that this may have anything to do with the issue but a peer has suggested that we define our duplex and speed on the nis and switch side- currently they are set to auto/auto.
