Problem with Starwind v8

[fanello]

Hi,

I have a big problem with the new starwind software and a HA-Device.
Sometimes, Starwind terminates the iSCSI Connection to one of the servers to syncronize a Device. After syncronization is finished, one of the 2 servers reconnects to the target, but the server on which the device was syncronized can't reconnect to the target. I have to restart the server, that the server can reconnect. It is always the localhost, who can not connect to its own target. The other server can connect after syncronization is finished.

Regards
Benny

[anton (staff)]

Catch situation and provide us a with a StarWind log for it. Also detailed description of your config (device sizes, FLAT or LSFS, now many StarWind nodes, how many partners for LU, network topology etc) is required.

[fanello]

Where can I find the logs?

[anton (staff)]

Logs folder for StarWind installation. Pick up the most recent one.

Where can I find the logs?

[fanello]

I made two screenshots:

starwind1.png (88.5 KiB) Viewed 14286 times

starwind2.png (82.95 KiB) Viewed 14286 times

[fanello]

starwind-20140603-172020.rar

(16.08 KiB) Downloaded 961 times

[anton (staff)]

OK, checking...

[Alex (staff)]

Benny,
I'm afraid, I have to ask you to send several latest files from /log directories of both HA nodes.
Please, also write approximate time when HA node disconnects the client initiator with following re-connection issues.

[fanello]

Hi,

the Logs are attached.

Node 2.rar

(196.94 KiB) Downloaded 1125 times

[fanello]

NOde 1:

Node 1.part03.rar

(113.81 KiB) Downloaded 1201 times

Node 1.part02.rar

(230 KiB) Downloaded 1106 times

Node 1.part01.rar

(230 KiB) Downloaded 1345 times

[Anatoly (staff)]

Hi!

Sorry for the long reply time.
we have reviewed the logs and here is what we get:
1.It looks like the unsynchronization happens because of the huge latency (>10s) on the physical storage where the StarWind images reside. I`d recommend you to take a closer look at it, since it may indicate the hardware problems.
2.Regarding the "reconnecting..." status of the targets in the MS initiator - we see that StarWind accepted the connections from following initiators:
iqn.1991-05.com.microsoft:hyperv-1.batterium.local
iqn.1991-05.com.microsoft:hyperv-3.batterium.local
Baasing on that fact we assume that this maybe something to do with the MS initiator, that can not refreshing the status for a long time. jsut doublecheck if hte disks are physically presented to the client machines.

Also, since you are a paid customer I will kindly offer you to contact our support by emailing to support@starwindsoftware.com and one of my colleagues will contact you. That should make the support process much faster.

I hope that helped.