[EMAIL PROTECTED] wrote: > I have linux machine and reo1000 iscsi storage. > > iscsi version: open-iscsi-2.0-865.9 > reo have latest firmware version 3.1.0.39 > > This machine works like file server, it means it hase 2 nics, one of > them is reo connected+reiserfs on it, the other one connected to local > network (it runs samba and nfs (for test) daemons). Local network nic > is 100 megabit, connection to reo is 1Gb. Transfer speed not very big, > cause it's old machine (PIII-800), no more than 3 MB per sec in case > of smb and upto 5 in case of nfs. > > If i starts writing on this or reading more than 300 MB of data, from > that machine in logs i saw: > > Apr 18 11:03:29 rat kernel: connection1:0: iscsi: detected conn error > (1011) > Apr 18 11:03:29 rat iscsid: Kernel reported iSCSI connection 1:0 error > (1011) state (3) > Apr 18 11:03:30 rat iscsid: received iferror -22 > Apr 18 11:03:30 rat last message repeated 2 times > Apr 18 11:03:30 rat iscsid: connection1:0 is operational after > recovery (1 attempts) > Apr 18 11:04:43 rat kernel: connection1:0: iscsi: detected conn error > (1011) > Apr 18 11:04:43 rat iscsid: Kernel reported iSCSI connection 1:0 error > (1011) state (3) > Apr 18 11:04:44 rat iscsid: received iferror -22 > Apr 18 11:04:44 rat last message repeated 2 times > Apr 18 11:04:44 rat iscsid: connection1:0 is operational after > recovery (1 attempts) > > In case of writing i can saw it more often. But file writes and reads > normally at all as i can see.
I think you might be trying to do too much IO. The default settings assume a decent target and connection. With only 100 mb ether we are probably trying to send too much IO to the box. Go and get open-iscsi-2.0-865.15. Use the kernel modules and tools from there and try these settings in /etc/iscsi/iscsi.conf: node.session.cmds_max = 4 node.session.queue_depth = 2 You can play around with those values and try increasing them too. But the default values of 128 and 32 are too high, and are probably going to cause IO to timeout and cause the error handler to run. > > I try to find any logs in reo, but it looks like all good. > > Also i try googling about "iscsid: received iferror -22" no info > about. You can ignore this. I am going to remove this error message. It just means that the kernel could not support a setting that userspace was trying to set. It is not fatal and the kernel can go a long without it. If it it was fatal then you would see some other value. > > Some one can tell me in which direction i must investigate problem? > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "open-iscsi" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/open-iscsi -~----------~----~----~----~------~----~------~--~---
