I have a windows 2012 R2 Server running it's iSCSI initiator to a readyNAS (516, tested with both 6.2.4 and 6.3.5RC2) target. A single LUN is created in a single group. All encryption/authentication is turned off.
I can initially connect just fine normally, but when I shutdown/reboot the Windows Server, the Windows server is unable to reconnect to the iSCSI target when it comes back up from the reboot. The iSCSI initiator shows "reconnecting" but is just stuck like that and it won't reconnect. Deleting and recreating the configuration on the Windows side does no good - Windows iSCSI initiator is unable to reconnect (or connect) to the iSCSI target.
If I reboot the NAS, it reconnects.
A bit less drastic than rebooting the NAS, I can also use the NAS web interface to pull the LUN out of the target group, destroy the now empty group, create a new group, and put the LUN into the newly created group. Then, suddenly, windows will reconnect to the iSCSI target.
Obviously, having to manually intervene any time I do an auto reboot of the server is a royal PITA.
The readynas interface doesn't really show any iscsi related stats, so I have no idea what it THINKS might be going on that would block a reconnect. Windows event logs are almost as useless, showing only that it failed to reconnect.
It's a 1TB "thin" provisioned, and this has occured ever since it was newly created (firmware 6.2.4) and empty. Currently, it's at about 29% used. (728.9 GB free space.) The LUN options are currently set as: BitRot is ON. Compression is OFF. Snapshot is OFF. Sync is ALLOWED.
What about if you instead use a thick LUN with bitrot protection and sync writes both disabled?
What is the capacity of your data volume and how full is this?
I'll have to wait for the weekend to try it with a completely different LUN configuration. The overall data volume is something like 6 or 8 TB, but it's only around 50% full. (So, I do have the room to make this LUN thick.)
Mdgm, are these suggestions along the lines of "likely solutions", or more like "trying different things to see what might work"?
The reason I'm asking is that I'm suspecting, based on other threads on here of a similar nature, that the problem I'm seeing is related to the iSCSI software on the NAS not releasing the connection when the Windows Server reboots. If that's the case, then the target might be refusing the connection after a reboot as it thinks that the initiator is ALREADY connected. Is there some way, from SSH, to see what's going on from the NAS point of view? (ie: for example, something that might show "current list of connections" on the target?)
Ideally, even if I work around the problem by using a thick LUN (assuming that works around the problem), I'd still like to understand WHY I'm having the problem to begin with so that, perhaps, netgear can fix it (or I can report the issue to Microsoft if it's with their initiator.)
The settings I mentioned (and all others how you have configured them of the ones you mentioned) are recommended settings and keeping the volume below 85% full.
Is iSCSI configured to automatically reconnect on your PC?
Is iSCSI configured to automatically reconnect on your PC?
Yes. As for opening a support ticket, I guess now I'm going to have to wait until the next official firmware release. (I'm running the 6.3.5RC2 right now - which was done as an attempt to resolve the issue. I somehow doubt that netgear will support the RC....)
