Weird disk sizes in Virtuall Iron

[gregg_hughes]

Good afternoon, all!

I am working my way through testing Open-e and Virtual Iron as our iSCSI solution. I have a problem that the VI people haven't yet been able to solve.

Running Open-e demo on SuperMicro 6025B-3 servers and VI 4.5.16 demo. Created two logical volumes: lv0000 is 300G File I/O and lv0001 is 150G Block I/O. I assigned them target0 and target1, respectively. My cunning plan is to run my virtual iron machines from file disks on lv0000 and point the SQL and Exchange databases to block disks carved out from lv0001.

When I try to assign disks in Virtual Iron and set up the iSCSI inside, I get some very strange values. iSCSI(1) shows 161.061G and iSCSI(2) shows 321.892G. Now, I can see some variance due to defining 1000 v. 1024, formatted vs. unformatted size, and so on, but I have some very real problems with this - not the least of which is I can't trust whether VI has identified the targets properly and assigned them in order and Identify the Block or File I/O devices!

I've rebuilt the disks twice in Open-e (a valuable learning experience!) and reloaded the VI to start it from scratch. And scratch is exactly what I'm doing to my head to figure out what's going on here.

Anyone got any ideas?

Thanks!


Gregg

[To-M]

I have not seen this issue before where the size is larger then what was created. Seeing that you tested with both File and Block IO I can't imagine how it would increase the values.

Does this happen if you test with just a Microsoft system using the MS iSCSI initiator?

[gregg_hughes]

The Microsoft iSCSI initiator showed the correct values - 299.78 and 150 Gigs respectively for the two volumes.

Any suggestions?

Thanks!


Gregg

[To-M]

Not sure why as well. Unless there is a Virtual Iron initiator setting that you can match with ours. You can go to the console and enter CTRL + ALT + W then Tuning options then iSCSI daemon options then select the Target and compare with Virtual Iron.

[enealDC]

Well I have 3 Open-E systems (2DSS and 1 ISCSIR3) deployed for VI and I've never seen this issue. It's truly weird.

In /opt/VirtualIron/VirtualizationManager/system/logs/node.ip.address.log - what are the reported sizes for the disks there?

Do you mind emailing me your logs file from Open-E and log files from running getSupportInfo.py on Virtual Iron?

[gregg_hughes]

How does Open-e calculate disk.volume size? In other words, how big is a gig?

Thanks!


Gregg

[To-M]

1 kilobyte is equal 1024 bytes and so on.