SUMMARY: ufsrestore problem (Tom Hart)

From: Tom Hart <tom.hart_at_huntingtonhospital.com>
Date: Fri Jul 19 2002 - 16:48:24 EDT
Thanks to Dave Mitchell for pointing me towards the solution to my
problem (duh!).

His suggestion helped me to determine that ufsrestore was spending an
extreme amount of time
Processing zero length files (that probably shouldn't have been there in
the first place). After stopping ufsrestore and re-starting it minus the
aforementioned files the restore completed in a couple of hours.


> I am running a Sun Netra with 2x300 Mhz procs, OS is 5.6. I have had
> to restore a 4.8 Gb filesystem to a raid 5 array due to the failure of

> 2 drives. The restore is running with no error messages displayed from

> a DLT7000 drive. Problem is this, the restore has been running for 32
> hours and has 200 Mb to go. Currently, the restore is proceeding at
> the rate of 10k every 5 minutes or so. The restore started out by
> writing an esitmated 60 to 80 percent of the data in the first hour or

> so, since then the rate has slowed consistently with the passage of
> time. The processor utilization has increased from an initial 2%
> slowly to 48% now. I have had to restore relatively small user
> directories in the past (100 - 500mb) that took some time (~1.5hr) but

> nothing like this.

Try running

    # truss -d -p NNN

where NNN is the process-id of the ufsrestore process,
to see where the process is spending most of its time (is it hanging
waiting for data from the tape, or ...)

Dave.
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
Received on Fri Jul 19 16:55:46 2002

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:42:49 EST