This post is more than 5 years old
2 Posts
1
1726
October 12th, 2010 11:00
Difference and impact between EMC "in and out of family" microcode upgrade
Hello,
I'd like to find out what's the difference between EMC in family microcode upgrade and out of family microcode upgrade on a DMX3/4 and V-Max
from one major code to another major code upgrade? i.e. from 5771 to 5773.
I know it's done online, but can it be disruptive or none? Is there any impact on production hosts and DR hosts during in family and out of family upgrade?
Should the hosts be powerd down or no? How long does an upgrade take normally? What if something goes wrong, is it easy to back-out and revert to the previous code?
Any explanation and additional information will be helpful and appreciated.
Thanks in advance.
No Events found!
dynamox
9 Legend
•
20.4K Posts
1
October 12th, 2010 11:00
i can share with you my experience with upgrading DMX3 from 72 to 73. First of all we had to collect and submit emcgrab/emcreport for every single box connected to that array. Our CE ran them through HEAT report (you can run HEAT reports yourself now) to verify that PowerPath, drivers, firmware were all up to par according to EMC interop matrix. CE did the upgrade during our lowest I/O periods, we did not notice a thing, PowerPath never blipped. Even our old OpenVMS boxes did not notice anything.
RRR
2 Intern
•
5.7K Posts
0
October 13th, 2010 01:00
I can only remember a 70-71 upgrade and that one was disruptive for 10 Windows servers (about 300 hosts in total, amongst which Windows, Linux, AIX, HPUX, Solaris). We also had to provide emcgrab outputs prior to the upgrade, but the hickup for these 10 was unforseen. I noticed however that from that moment on the HEAT output displayed an extra warning about hosts being certified for a Nondisruptive Upgrade or not. My guess has always been, that I triggered the extra line in the HEAT output, that you are now looking for
Booyah2
2 Intern
•
185 Posts
2
October 13th, 2010 07:00
If done properly, it is Non-Disruptive. But the GRABs are very important as well as having your local EMC CE and FSS support team investigate your environment and NDU alerts, which is part of the EMC change control process.
As an example of things you may run into, depending on your host environment:
emc224025, emc173690, emc203233, emc121956, emc212937.
Hope this helps...
Note: if you find any of the replies to your question as helpful or answering fully, please be so kind as to mark replies appropriately.
SANshine1
2 Posts
0
October 13th, 2010 11:00
Thank you all for your prompt and very helpful answers. Your replies helped me better understand the process and be prepared for my next enginuity upgrade.
SANshine