Actually, it downloads the FW to the shelf, then the shelf will copy it to both ESH or LRC or whatever modules.
So, yes, even in a cluster, upgrading the shelf firmware from one head should take care of all shelves...as long as that one head can see all the shelves on at least one loop.
--tmac
On 3/9/07, Nils Vogels bacardicoke@gmail.com wrote:
On 3/8/07, tmac tmacmd@gmail.com wrote:
Personally, I would:
copy the 7.1.2_setup_x.exe file to the /etc/software directory on each filer.
Then from the console of each filer:
options raid.background_disk_fw_update.enable off (turns off auto updating of disk firmware) software install filename (the filename from above)
While at it, get and extract the latest all.zip and all_fw.zip files (disk and shelf FW) into the /etc dir on each filer.
Optionally, with cluster still enabled, run disk_fw_update to update all disk firmware (lessens downtime at reboot) usually only needed on one head. When complete, run on other head to be sure. storage download shelf to update shelf firmware (lessens downtime at reboot) usually only needed on one head. When complete, run on other head to be sure. options raid.background_disk_fw_update.enable on (turns the auto updating of disk firmware back on)
I'm preparing to do a NDU 7.0 -> 7.1 and was wondering: When you do the storage download shelf on one head, will it update the firmware on both the A-loop and B-loop ?
I would expect not, since one head is only connected to one loop in a plain cluster config. AFAIK, downloading shelf firmware takes a quiesce of a loop and a controller reboot, and my customer really would like to have a functional system during this NDU.
Greets,
Nils
-- Simple guidelines to happiness: Work like you don't need the money, Love like your heart has never been broken and Dance like no one can see you.