All of a sudden, we are not able to deploy from a template to one of our ESXi 4.1 U3 clusters. It goes to 87% and then we get "Could not complete network copy for file". We can deploy from this same template to other clusters. I SSH'ed to the host we happen to be deploying to and noticed the following errors get logged in the messages log as it's happening.
Feb 8 17:18:48 vmkernel: .... cpu2:5469)BC: 2591: Blocking due to no free buffers. nDirty = 1764 nWaiters = 3
Anyone seen this before? I've opened a SR.
Also, I should mention, though it could be unrelated - We've had a few strange HA issues that have affected at least this cluster and another cluster intermittently, where the HA Agent thinks it can't talk to other slaves/master, though if you manually ping from the host as it's happening the replies come back just fine. When this happens, the logs spew messages such as "...XXX.XXX.XXX.XXX (IPs of other hosts in cluster) is bad IP", etc.