Unsolved
This post is more than 5 years old
33 Posts
0
561
NS120 getting code upgrade:impact to ISCSI Vmware?-increase vm timeouts
I was wondering what others have seen. The data movers will obviously reboot. If it takes say 3 minutes, I beliveve VM's will start to error out. Have admins dialed up ESX iscsi timeouts so it can 'just wait' for the data mover failover?
Anything else I can expect? we mostly use the NS for virtual servers- as well as a few direct attached.
I am going from 5.6.44-4 to the latest approve 6.x.x.x
dynamox
1 Rookie
1 Rookie
•
20.4K Posts
0
January 12th, 2011 08:00
as far as i remember, out-of-family upgrades required that both primary and standby datamovers reboot at the same time, so that time period where storage is not available might be a lot longer than 3 minutes. Maybe somebody who's gone through this upgrade can share their experience.
umichklewis_ac7b91
300 Posts
0
January 12th, 2011 13:00
Just went to DART 6.0 on December 28th and our NS80's primary/standby datamovers were down during the reboot. Both DMs were down for 7 minutes and 30 seconds from last ping, so I can imagine iSCSI would die a little earlier.
My advice is to negotiate for a scheduled downtime window to do the ugprade. If you tell them 10 minutes and it suddenly takes 12 minutes, you'll probably irritate someone.
Good luck!
Karl
pvega1
2 Posts
0
January 18th, 2011 09:00
I have been through two code upgrades on an ns-20 and ns-120 and things have gone wrong. I lost a few vm's because a few vmfs volumes were corrupted while doing it live. Do yourself a favor and schedule downtime.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
January 20th, 2011 12:00
I know it's a big read, but the following guide is available on Powerlink and tells you everything you need to know as far as configuring your VMware environment to withstand Celerra data mover failovers: