Has anyone successfully used Direct SAN Access to an IBM DS3500 with dual controllers? If the LUN has controller B as preferred when the backup job starts a Critical Alert is sent from the DS3500 with the following description "Logical Drive not on preferred path due to ADT/RDAC failover" and the LUN is now on controller A; the job continues using direct access and everything looks great except for that alert.
Here's my setup:
Physical box running Veeam B&R 6.1
2 NICs - 1 leg in Admin LAN the other in the iSCSI LAN (Separate LAN from Admin using 10.0.0.X range)
IBM mpio client installed on server
MS iSCSI initiator configured for discovery to 10.0.0.1, which shows all volumes on both controllers as Basic/Offline Healthy(Primary Partition)
DS3500 - With LUNs split across controllers.
Controller A - IP range 10.0.0.1 - 10.0.0.4
Controller B - IP range 10.0.0.5 - 10.0.0.8
I wonder if it is a zoning issue caused by both controller ip ranges residing in the same 10.0.0.X.
Any help is greatly appreciated.
-
- Lurker
- Posts: 2
- Liked: never
- Joined: Feb 15, 2011 4:19 pm
- Full Name: Chris Cable
- Contact:
-
- Lurker
- Posts: 2
- Liked: never
- Joined: Jul 10, 2012 10:33 pm
- Contact:
Re: B&R 6.1 Direct SAN Access to IBM DS3500 SAN
We had this issue with a DS3300 (not with backups, just with reboots, etc). The LUN's would error and say that they are not on the preferred path. We updated the firmware on the SAN (be careful with this BTW, you need to subscribe to the alert emails on firmware revisions... I HATE IBM SANS!!!!) and changed the preferred path to the other controller.
Remember, this is just a warning, not an error. I think it has something to do with MPIO. If you backup server is preferring the opposite path that the SAN is preferred for, the SAN will move paths on you.
Good luck,
Linuxx
Remember, this is just a warning, not an error. I think it has something to do with MPIO. If you backup server is preferring the opposite path that the SAN is preferred for, the SAN will move paths on you.
Good luck,
Linuxx
-
- Veeam ProPartner
- Posts: 27
- Liked: 4 times
- Joined: Jan 31, 2012 2:00 pm
- Full Name: Giorgio Colucci
- Location: Italy
- Contact:
Re: B&R 6.1 Direct SAN Access to IBM DS3500 SAN
Hi all,
do you have an update for this issue?
Do you know if it's possible to map the LUN to a Windows Server (Veeam Repository) in READ-ONLY mode (just to prevent some LUN distruption)?
Thank you
Giorgio
do you have an update for this issue?
Do you know if it's possible to map the LUN to a Windows Server (Veeam Repository) in READ-ONLY mode (just to prevent some LUN distruption)?
Thank you
Giorgio
Giorgio
-
- Lurker
- Posts: 2
- Liked: never
- Joined: Feb 15, 2011 4:19 pm
- Full Name: Chris Cable
- Contact:
Re: B&R 6.1 Direct SAN Access to IBM DS3500 SAN
Giorgio,
As far as I know the IBM DS series SANS do not have the ability to set read-only access.
Chris
As far as I know the IBM DS series SANS do not have the ability to set read-only access.
Chris
-
- Expert
- Posts: 230
- Liked: 41 times
- Joined: Feb 18, 2011 5:01 pm
- Contact:
Re: B&R 6.1 Direct SAN Access to IBM DS3500 SAN
I know this is an old thread, and I'm not sure if this will fix the issue, but the IP address you have configured for your two controllers are not the IBM recommended configuration. You're supposed to each of the four ports on the controller have an IP in a separate network. Since each DS3500 controller has four ports you'd have four different networks and each host would have four separate VMKernel iSCSI ports, one for each network. So, for your configuration it might look something like this, assuming your mask is 255.255.255.0:chrisc wrote:Has anyone successfully used Direct SAN Access to an IBM DS3500 with dual controllers? If the LUN has controller B as preferred when the backup job starts a Critical Alert is sent from the DS3500 with the following description "Logical Drive not on preferred path due to ADT/RDAC failover" and the LUN is now on controller A; the job continues using direct access and everything looks great except for that alert.
Here's my setup:
Physical box running Veeam B&R 6.1
2 NICs - 1 leg in Admin LAN the other in the iSCSI LAN (Separate LAN from Admin using 10.0.0.X range)
IBM mpio client installed on server
MS iSCSI initiator configured for discovery to 10.0.0.1, which shows all volumes on both controllers as Basic/Offline Healthy(Primary Partition)
DS3500 - With LUNs split across controllers.
Controller A - IP range 10.0.0.1 - 10.0.0.4
Controller B - IP range 10.0.0.5 - 10.0.0.8
I wonder if it is a zoning issue caused by both controller ip ranges residing in the same 10.0.0.X.
Any help is greatly appreciated.
Controller A
Port 1 - 10.0.1.1
Port 2 - 10.0.2.1
Port 3 - 10.0.3.1
Port 4 - 10.0.4.1
Controller B
Port 1 - 10.0.1.2
Port 2 - 10.0.2.2
Port 3 - 10.0.3.2
Port 4 - 10.0.4.2
Who is online
Users browsing this forum: No registered users and 36 guests