Unsolved
This post is more than 5 years old
1 Message
0
14942
(2) 6224 Stacked switches in VMWare environment, Failover not working!! Please help
We have recently implemented a new VMWare 4.0 ESXi environment and ISCI Storage. We are currently in testing mode. We have 2 6224 Powerconnect switches stacked for our production network side of VM and 2 stacked 6224 switches for our ISCI network which those switches are completely segmented off the production network..
We originally had the issue we the master switch went down on the network production side the slave would do a restart and come back as Master, during that process all of our VM's would crash. Now with the new 3.2 update the slave switch does a very quick warm restart and reverts to the new master switch. No downtown.
The issue isn't pulling the power of the master switch and the slave taking over, everything stays up, its powering back up the old master switch and as it boots up into slave mode I lose connection and all my VM machines crash.
I've had STP enabled, disabled, switched around my network cables and how they were plugged into the switches, always the same result.
On the ISCI switch side I have something similiar happening except just powering down the original master I loose my ISCSI network connections, but as the slave becomes the master and I power back up the old master into slave mode, everything comes back. Strange thing is when I power down the newly promoted master I get expected results and I only loose half my connections and my VM's don't crash.
There seems to be some kind of strange behavior with the ORIGINAL masters in stacked configuration.
Hope this makes sense and hopefully someone can help me out, I need complete redundancy and I might have to look out spending some serious dough and moving to Cisco switches, I haven't heard of any issues like this from Cisco.
Chris Slack
<ADMIN NOTE:Email Id removed per privacy policy>
Vinny79
1 Message
0
January 14th, 2011 05:00
Hello Chris,
I got the exact same problem with the exact same switches. I have this problem with Equallogic FW 4.3.6 and 5.0.2. I also have the problem with ESX versions 4.0 and 4.1.
It seems that failover in one direction is OK, but once I failover into the other direction, I've got much longer time-outs. The time-outs aren't big enough to cause problems on the VM's Windows OS.
But it must be said that this is a brand new environment and I'm now testing on clean OS'es without applications.
When testing it seems that the time-out sometimes exceeds 60 seconds, but that is just a guess.
What I wondered is if you did solve this problem?
Kind Regards,
Vincent
PS: Fyi: I used these docs to configure the environment:
"Configuring VMware vSphere Software iSCSI with Dell EqualLogic PS Series Storage"
"Dell Equallogic configuration Guide"
"ISCSI SAN COnfiguration Guide" from VMWare (vsp_40_iscsi_san_cfg.pdf)