Printable View
All,
I am trying to use storage motion to move a running vms disks to an open-e server FC Lun but I am getting a time out problem and a critical error in the open-e hardware system_load log as shown below
03-02 21:25:53.344699 CRITICAL System overload!! Avg WA(Waiting for I/O operations): 27.34
gajh
these tweaks will help:
http://kb.open-e.com/How-can-I-fix-C...ssage_202.html
make sure to be on the most recent release of DSS.
making these tweaks will reset target connections.
In failover configuration, both nodes must have the same target values, before failing over.
Thanks for the reply so this applies to Fibre Channel Luns as well as iscsi?
if i reboot the sever then it works for a while then errors again
How much RAM in the machine ?
16GB Memory, it works ok with iscsi but FC seems to cause the problem
It also causes the admin gui to dropout
The ESX server error reports "a general system error occured: Source detected that the destination failed to resume"
check the raw disk speeds in the test.log file from the log package in DSS.
Some performance tweaks may be needed. specifically around caching.
Also check the firmware on the FC card to make sure its current, as well as proper slot for the speeds.
Hi I have downloaded the logs looked at them but it means nothing to me to be honest
I have upgrade all firmware. When I move a vm to the a fc volume on opene the esx hosts keeps getting a "lost access to volume due to connectivity issues" error and then recovers, the speed is ok but I dont get errors when i move the same vm to a different vendors storage server.
send me the dmesg logs(via PM) and I can have a look.
the 2 you sent look ok for everything on boot.
anyway you can send me the entire package ?
Logs sent, also another thing that has happen last night I was coping a large file to an iscsi volume and the server stopped responding, no ping, console was frozen couldnt do anything anything had to do a power off to get it back on. I havent looked at these logs yet
Quote:
Originally Posted by Gr-R
are you still getting high load messages?
It looks as though the discs were initializing at the time you saw them before.
block I/O will probably give you better performance with the LUNs as well.
No the load messages have gone but I am still getting the disconnects from the LUNs, I created an iscsi LUN and this was working then the open-e froze while doing a 70GB file copy of a vm machine, I had to reboot the open-e server to recover
make the target adjustments mentioned here:
http://kb.open-e.com/How-can-I-fix-C...ssage_202.html