Hello,
I had intention to use ZFS with mirrored pools on external HDDs.
Unfortunatelly, I have some troubles to have it running as planned.
I''m enclosing my results so far. Please let me know, if some of
you have similar issues or not. Thank you.
Firstly, I can''t access 1TB disks via FireWare. Is someone using
IEE1394 connection?
Secondly, I have two sets of external disks for USB connection.
1. 3x 1TB disks in external cases from Seagate, Maxtor ad WD. This
connected to active USB hub and then to notebook / PC.
2. 3x 1.5TB disks inserted in cases with SATA/USB coverters, this to
active HUB and this to notebook/PC
32bit snv_108 is reporting, when only one 1.5TB connected, that
disk has 2930277168 blocks, which is too large for a 32-bit kernel
I was able to rsync my data to these pools (and verified via rsync) But
when the pool is used, soon I''m getting UNAVAIL / DEGRADED status on
these disks. I noticed, that on faster machine, the issue is comming
later.
for example zpool status extpool
svazek: extpool
stav: DEGRADOVAN?
status: One or more devices are faulted in response to IO failures.
action: Make sure the affected devices are connected, then run ''zpool
clear''.
viz: http://www.sun.com/msg/ZFS-8000-HC
operace scrub: resilver in progress for 8h7m, 12,03% done, 59h27m to go
nastaven?:
NAME STATE READ WRITE CKSUM
extpool DEGRADOVAN? 30 0 0
mirror DEGRADOVAN? 68 5 0
c10t0d0 UNAVAIL 59 1,32M 0 nelze otev??t
c9t0d0 UNAVAIL 33 1,31M 0 nelze otev??t
c11t0d0 ONLINE 28 5 0 2,71G resilvered
chyby: 2 chyby dat, pou?ijte p?ep?na? ''-v'' pro seznam
(sorry for czech, export LANG=POSIX is ignored)
I think, that my issues are related to size of disks. My mirror of
32bit notebook''s rpool to external USB 160GB is working just fine.
I also noticed, that on faster machine the issues come later.
scrub can''t be run. I''m getting quickly all disks UNAVAILABLE
I wish to understand, what I can do, to have this working.
Thank you,
Jiri
--
Jiri Navratil, http://www.navratil.cz, +420 777 224 245