Dual LVM path configuration verification prior to AutoRAID Disk Array 12 battery replacement

[Christopher Caldwell · Accepted solution]

We recently experienced a similar problem. The CE replaced the battery on the fly - it takes about 24 hours for the new battery to show normal.

We built connections to the Autoraid using stripes (one LUN through control X, one LUN through controller Y, LV striped across both controllers).

On vgdisplay -v of the AutoRAID vgs, you should see something like
PV Name /dev/dsk/c3t0d0
PV Name /dev/dsk/c4t1d0 Alternate Link
for all of the PVs (LUNs). If you see the dual paths, you have redundancy.

You can cross the paths with what's in lvdisplay - you should see something like:

PV Name LE on PV PE on PV
/dev/dsk/c3t0d0 469 469
/dev/dsk/c4t1d1 469 469


Risks of changing on the fly are - dataloss because something goes quite wrong. I don't think that scenario is likely - the CE seemed quite comfortable doing this hot, but it's a possibility.

Therefore ...
make sure you have two good backups prior to the replacement.

We did the swap at a low traffic time of day, so the controllers weren't too busy.

[Eugeny Brychkov]

Strongly agree! Make sure before you'll do anything to Autoraid that you have GOOD backup of ALL data. In addition to Christopher's reply I would pay your attention to your Autoraid having very old firmware. As I remember there is HP62 firmware. If possible let HP to do this job on firmware upgrade and battery replacement
Eugeny

[Yogeeraj_1]

Dear all,

thank you for your comments.

We will be doing the replacement of batteries tonight.

It has been decided that we would be doing "the Online dual controller battery replacement" but with the server down! We don't want to take any risks...

We will not be doing any firmware upgrades since we will be soon depreciating the old k250 server and the attached 24GB AutoRAID at the expense of our 2 new RP5430s and 100GB LSI STK D173 storage array!


Best Regards
Yogeeraj