What OS is the Samba share on?
It's kind of odd, if I ctrl-c out of bmi when it first comes up (right after it starts and just sits and waits for instructions from the provisioning server, basically where it has the .............. printed out on the screen), I can execute a manual net use command just fine and it mounts the windows share and I can cd into it and do a dir and it feels fast, no delays. But if I ctrl-c after I give the bare metal server the system package to provision, and I cd into the k: drive (so I let the provisioning process automatically mount the drive), it all of the sudden is slow (the k: drive) and takes forever to return when doing a dir or the like.
If I mount the same exact network share on my laptop and try to transfer the same I386 directory back to my laptop, the transfer starts but it says its taking 1 hr to fully transfer it. Not too sure if its fast or slow, but feels slow to me.
On the bare metal server, if I ctrl-c out, I can ping the datastore just fine and it has a fast response time.
I also tried specifying the hostname and then the ip as well in the datastore property dictionary and no differences.
Has anyone ever done any special configuration on the switch directly on the hs21 bladecenter chassis ?
Within winpe just doing a copy command in the cmd window, like this:
copy k:\win2k\i386\adminpak.msi c:\
it says "the specified network name is not found" still.
What's odd is using the same winpe image for a different bladecenter, it works fine, the full provisioning process, not just the copy command.
Have you tried a different blade in the same bladecenter? Sounds like it could be a chassis config or integrated switch config issue.
I would try the known good HS21 in the exact same slot as the one giving you issues. I have seen H chassis's with bad slots/backplanes where nic and or FC connections were defective ether on the blade side or the backplane side.
It was odd, we figured it out, it was because the win2k app server nic card wasn't set to 100 mbps, it was set to auto, and it really didn't like that with their cisco switches. The other server that was working was on a different switch. When we forced the app server to 100 mbps, then it worked much much better.