On Fri, Feb 20, 2009 at 05:07:49PM +0100, Ulrich Windl wrote: > > On 20 Feb 2009 at 10:19, Konrad Rzeszutek wrote: > > > > > > Maybe that's related: With SLES10 SP1 (x86_64) I have two iSCSI LUNs that > > > are > > > reachable over 16 paths each, and the networ connections use two > > > dedicated > > > separate VLANs. As it's a test at the moment, there's only one initiator > > > and two > > > iSCSCI gateways connected to that switch. And there is no traffic (the > > > filesystem > > > is a RAID1 that is not mounted). Still I'm seeing message like this: > > > > > > Feb 19 01:39:50 rkdvmso1 iscsid: connection198:0 is operational after > > > recovery ( > > > 1 attempts) > > > Feb 19 07:49:30 rkdvmso1 iscsid: connection215:0 is operational after > > > recovery ( > > > 1 attempts) > > > Feb 19 08:53:15 rkdvmso1 iscsid: connection208:0 is operational after > > > recovery ( > > > 1 attempts) > > > Feb 19 14:09:40 rkdvmso1 iscsid: connection211:0 is operational after > > > recovery ( > > > 1 attempts) > > > Feb 19 19:33:42 rkdvmso1 iscsid: connection220:0 is operational after > > > recovery ( > > > 1 attempts) > > > Feb 20 01:43:56 rkdvmso1 iscsid: connection210:0 is operational after > > > recovery ( > > > 1 attempts) > > > > It looks as if each path gets bumped. Is there anything related in the > > target > > around those times? Did you see a pattern over a long time of which path > > and what > > time they are being bumped? > > > > Also, are there any backup being done over that time-frame that would > > saturate > > the switches? > > Hi! > > I'll have to investigate a bit deeper. If it's not the network, it could be > FC or > the target hardware itself. There are some backups actually running, some of > those
Those are iSCSI errors. How does FC figure in this? Is FC used for backups? If so, then that shouldn't be an issue as the FC and network fabric are seperate. > I'm responsible for, and others I'm not responsible for, but they use the > same > infrastructure, and even others that only use the FC ISL (Inter-Switch-Links) > for > data transport. However we were believing that FC guarantees maximum delays > and no > packet drops, so I dobt that it's the FC, but who knows? I'll have a deeper > look, You will get errors when the FC packets get dropped and can't be re-sent. > comparing it to the backup times I know... > > Regards, > Ulrich > > > > > > > > > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "open-iscsi" group. To post to this group, send email to open-iscsi@googlegroups.com To unsubscribe from this group, send email to open-iscsi+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/open-iscsi -~----------~----~----~----~------~----~------~--~---
