Solaris

do you mean:

Code:

# metareplace d100 c1t0d0s5 c1t0d0s5

in this case system gave error:
metareplace: sspfs_svr: c1t0d0s5: has appeared more than once in the specification of d101

I think got it now:

Code:

# metareplace -e d100 c1t0d0s5

Can metastat -e be done online . How much is the load it puts on server resources .

We had the device /d45 coming in 'needs maintenance ' status about a month ago . I c the same disk coming in maintenance today. Does this mean I need to replace the disk.

Did that fix your issue?

You will usually know immediately if it fails. Sometimes it will sync up almost all the way before crapping out.

I always do a "metastat|grep %" to monitor the sync progress.

also, another thing I will do is do a "metastat|grep Invoke" to see how many slices are hosed.

metareplace -e <mirror> <device> (metreplace enable command) worked in case of disk needs really resynchronisation. I think it meant like Sun Solstice DiskSuite said is useful when a component fails due to human error (for example, accidentally turning off a disk)..

But some of the disks needed physical replacement even though metareplace -e command cleared the errors in the first place.

Also consider the commands; iostat -eE or iostat -en result to see HW errors..

after we replaces a disk with metareplace -e option..,

here do we need to resync ??. how the new disk will update with the old data.

please advice.., thanks