I did. I setup 192.168.0.120 and 192.168.2.120. And deleted all bonds.
But I get lots of data_out_start(1302) messages from open-e. And system goes to read-only mode.
My current configuration is:
2 servers are running CentOS5.2 + GFS + VMWare Server.
Boot from iSCSI and GFS is on iSCSI.
also, using qlogic iscsi HBA.
Also, when I'm using MPIO, when ever a single server's iowait is over 40%, CentOS 5.2 puts me in the read-only mode. And Open-E does not respond at all. Even from console, Open-E does not take any key stroke.
The data_out_start(1302) messages could be from different settings from the initiator and target.
Check to make sure that they match. From the iSCSI-R3 server go to the Console screen and enter CTRL + ALT + W then select Tuning Options then iSCSI daemon options select option 1 and choose target and compare the setting to match from the initiator.
Some claim that these are the preferred settings when working with VMware:
Concerning problems with the console not performing correctly, this could be a hardware issue but hard to tell without logs. Run a memory check, to do this restart the system and after POST enter the "Escape or TAB" key and select MEMtest. This will take some time (usually wait for 1 hour) but I have seen this before in some cases.
If you are still receiving these errors please send logs to support@open-e.com to have an engineer review them.
I think this slow performance kills my server. I tried given parameters.
Still I have slow performance and during the heavy I/O, vmware server timed out and my host OS cannot communicate to VMware. (when MPIO was off) When MPIO was on, entire host OS goes to read-only mode.