Recent content by julianbarnett

Thanks for those replies - you are right I need jukebox sharing. For flexibility we have zoned all drive paths to both the servers and storage nodes but then configured Networker to use the desired path via the rmt devices. In other words with 2 HBAs per server, Networker inquire can see 20...

Thanks for that. I have configured the last 2 drives in the jukebox solely with AIX device paths - will see how it runs tonight. I think the DDS option confused me - I thought it was the only way a libraray could be shared between server and storage node.

Hi - we have a SAN connected Scalar I2K with 10 LTO4 drives attached to a CPU bound Solaris Networker server. I tried using DDS to share out some of the drives to a SAN connected AIX storage Node and alleviate the load on the Solaris server. However, using this configuration, if the server is...

The bootstrap records are saved to tape - the notification just tells you where the most recent bootstraps are - tape and file position on tape. Imagine your networker server has a disk crash and you lose the entire /nsr partition. In order to recover you need to reinstall the networker software...

I believe you just need to change the Action in Bootstrap Notification - this info is important for Disaster Recovery so you may prefer to email it - on Unix we use the Action -/usr/ucb/mail -s "Bootstrap for <backup_server>" <mail.address@your.domain>.

savegrp -R <group_name> This option is used to restart a group that was stopped or if savesets failed and they need to be retried. The restart window attribute of the group is used to determine if it is too late to be restarted. If the window has elapsed the restart is converted into fresh start.

You will see this behaviour if the client in question is a Windows Cluster virtual resource - it uses the client resource of the underlying physical server. We have lots of these - e.g. db204a (physical server A), db204b (physical server B) and db204v (Virtual IP resource with shared disks). The...

oops - missed final ' ! mminfo -m -q "pool=<POOL>,full,!volrecycle" | grep -v volume | awk '{ print $1 }'

The Group Restart function in NMC is supposed to address this - as long as the group restart window has not been exceeded, only the failed saveset will be retried. We have updated all our daily groups to have a restart window of 23:59 to allow reruns to be perfromed up to 24 hours after initial...

mminfo -m -q "pool=<POOL>,full,!volrecycle" | grep -v volume | awk '{ print $1 } Should do the trick

I would apply patch, power cycle L9 and reboot system - should clear the error and reset SCSI channel to Fast/Wide. If it still runs slowly and comes up with SCSI errors get then it's time to get the L9 and SCSI card looked at.

Wonder if this is ufsdump or the filsystem causing the problem, can you make a large file in the target partition? e.g. mkfile -v 4000m /devbackup/largefile.

It's not volume manager - it's the filesystem you have mounted , the mount command will show largefiles against filesystems that are largefile enabled - I believe the standard ufs mkfs and mount options assume applications are largefile aware. Veritas I think assumes nolargefiles by default and...

P.S. What version of Solaris are you running? - check to see if you have the latest glm driver patch applied. This is 109885-16 for Solaris 8 - 109885-12 has some fixes for phase parity problems. Regards Julian

OK so all drives are Fibre attached and the only thing on the SCSI channel is the L9. From the glm driver man page.. Target <id> reverting to async. mode A second data transfer hang was detected for this target. The driver attempts to eliminate this problem by reducing the data transfer...