I have a server with a huge number of datasets (around 9000)
When the pool containing the datasets is imported on boot up, a few (<10)
datasets are not mounted and thus not exported via nfs. Which dataset is not
mounted is random.
All datasets are exported via nfs. A zfs import takes around 30 mins.
The system is running s10u4
sunsystem9:[~] > uname -a
SunOS sunsystem9 5.10 Generic_127112-10 i86pc i386 i86pc
The zpool configuration is as follows:
sunsystem9:[~] > zpool status
pool: private
state: ONLINE
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
private ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c6t600A0B800029A240000007204753ACF1d0 ONLINE 0 0 0
c6t600A0B800029A240000007224753AE8Fd0 ONLINE 0 0 0
c6t600A0B800029F7C4000007304753B2A2d0 ONLINE 0 0 0
c6t600A0B800029A240000007284753B6D3d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c6t600A0B800029F7C40000072A4753ACB4d0 ONLINE 0 0 0
c6t600A0B800029F7C40000072E4753B126d0 ONLINE 0 0 0
c6t600A0B800029A240000007264753B341d0 ONLINE 0 0 0
c6t600A0B800029F7C4000007344753BF84d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c6t600A0B800029F7C40000072C4753AE42d0 ONLINE 0 0 0
c6t600A0B800029A240000007244753B20Dd0 ONLINE 0 0 0
c6t600A0B800029F7C4000007324753B6BAd0 ONLINE 0 0 0
c6t600A0B800029A2400000072A4753C06Dd0 ONLINE 0 0 0
errors: No known data errors
The disks are on a ST6140 attached to the system via redundant fc (multipathing)
(Each of the disk devices are volumes with 5 500gb disks in a raid5)
Any ideas?
This message posted from opensolaris.org