EVA RSS reconstruct failed

[Greybeard]

Hi all, anyone seen this before?

An EVA4400 running 10000000 F/W with 72 x 450GB disks had a single disk failure, nothing in the logs to suggest any other issue but the remaining 7 disks in the RSS that lost the disk now have far lower occupancy per disk than all the others, and after replacing the failed disk it has been showing 0% reconstruct for days.

I know there have been issues with 450GB disk firmware in the past and this system is due some updates but even so, this is a bit odd.

Also for some reason the site guys put the replacement disk in a different bay, same shelf, but EVA perf shows the old disk (no I/O) whereas Command View and the config dump does not.

I'm hoping to schedule a reboot to see if a rescan of the hardware map helps but it's a 24/7 system :(

[tilba16narooma]

Dis you group the disk on replacement ?

[Greybeard]

Yep, and dropped protection level from double to single to ensure plenty of unallocated space. The disk group is at about 74% allocated.

[Greybeard · Accepted solution]

well "all" it took was deleting some data to gain space, remove all disks in that RSS from the group and a cold boot, then all F/W was patched once the group was behaving normally.