Hi,
I'm having some problems getting a PE2950 connect to MD3000i. The software adapter lists no targets (vm-iscsi.jpg), and I have tried rescaning several times even restarting twice.
ESX Server is version 3.50 82663 (Diden't work before upgrade to update 1)
Firewall is opened to allow Software iSCSI Client 3260(TCP).
/var/log/vmkernel shows:
Apr 12 15:23:18 vmkernel: 0:00:20:07.421 cpu5:1041)SCSI: 861: GetInfo for adapter vmhba32, , max_vports=0, vports_inuse=0, linktype=0, state=0, failreason=0, rv=-1, sts=bad001f
Apr 12 15:23:18 vmkernel: 0:00:20:07.424 cpu5:1041)WARNING: SCSI: 279: SCSI device type 0xd is not supported. Cannot create target vmhba0:288:0
Apr 12 15:23:18 vmkernel: 0:00:20:07.424 cpu5:1041)WARNING: SCSI: 1249: LegacyMP Plugin could not claim path: vmhba0:288:0. Not supported
Apr 12 15:23:18 vmkernel: 0:00:20:07.424 cpu5:1041)WARNING: ScsiPath: 3187: Plugin 'legacyMP' had an error (Not supported) while claiming path 'vmhba0:C0:T288:L0'.Skipping the path.
I can ping all four MD3000i iscsi host ports form the esx server (192.168.130.101, 192.168.130.102, 192.168.131.101, 192.168.131.102). ESX Network configuration can be viewed in the attachment (vm-network.jpg). The physical network is configured as shown in attachment (physical-network.jpg) with no vlan.
The iSCSI Initiator (vmhba32) is set to Dynamic Discovery on 192.168.130.101:3260 and 192.168.131.101:326 with no CHAP on either side.
Any help/suggestions would be greatly appreciated!
What is name of your iscsi initiator?
I can see iqn-
the - is not correct at that place.
The iqn must have the right format (with dot i.e iqn.2000-05.xxx:xxx check there:
http://wikis.sun.com/display/StorageDev/IQNNameFormat )
This is a Dell issue...there is a rather long thread already posted here which describes the problem and the solution. Search for "md3000i" and you'll find it. Basically, the MD3000i doesn't support the iSCSI standard for naming, so you have to go back in and make sure the name is less than 31 characters.
Indeed - this is part of why its not certified.
--Matt
Thanks for the reply. I changed the name to 18 characters on MD3000i, and esx server, but still no luck. Same error messages!
Well that is definitely supported with Esx 3.5.
Why don't you want to start with only one path to md3000i - and check the right fw version on your md3000i to run with Esx (I don't know the version no.).
Should be the both Gbit switches connected together? I'm not sure here.
I removed one of the vswitch, and took the physical cabling down to one path. I now have one pingable path down to the MD3000i. I also upgraded to the newest firmware version for the MD3000i.
Still doesn't work.
Just to clear something: Does the ethernet management ports need to be on my iSCSI network? I got both connected to my lan for management.
I saw that post earlier, and checked for the common mistakes. The LUN is around 1TB (which is total disk space in the md). All my initiator names or aliases are around 20 characters(less then 30), and the firewall is open for iscsi.
I ran the vmkiscsid in debug mode:
1208109712.195036 >> iscsid[3088]: Connected to Discovery Address 192.168.130.101
1208109712.195095 >> iscsid[3088]: starting iSCSI login on fd 7
1208109712.195143 >> iscsid[3088]: sending login PDU with current stage 1, next stage 3, transit 0x80, isid 0x00023d000001
1208109712.195169 >> iscsid[3088]: > InitiatorName=iqn-vmware-kyb-vm03
1208109712.195192 >> iscsid[3088]: > InitiatorAlias=kyb-vm03.FQDN (20chars)
1208109712.195250 >> iscsid[3088]: > SessionType=Discovery
1208109712.195295 >> iscsid[3088]: > HeaderDigest=None
1208109712.195329 >> iscsid[3088]: > DataDigest=None
1208109712.195363 >> iscsid[3088]: > MaxRecvDataSegmentLength=8192
1208109712.195386 >> iscsid[3088]: > DefaultTime2Wait=0
1208109712.195407 >> iscsid[3088]: > DefaultTime2Retain=0
1208109712.195428 >> iscsid[3088]: > ErrorRecoveryLevel=0
1208109712.195450 >> iscsid[3088]: > IFMarker=No
1208109712.195478 >> iscsid[3088]: > OFMarker=No
1208109712.195516 >> iscsid[3088]: wrote 48 bytes of PDU header
1208109712.195552 >> iscsid[3088]: wrote 232 bytes of PDU data
1208109712.203432 >> iscsid[3088]: read 48 bytes of PDU header
1208109712.203480 >> iscsid[3088]: read 48 PDU header bytes, opcode 0x23, dlength 0, data 0x9a097b8, max 32768
1208109712.203521 >> iscsid[3088]: discovery login to 192.168.130.101 rejected: initiator error (02/00), non-retryable, giving up
1208109712.203590 >> iscsid[3088]: disconnecting session 0x9a07780, fd 7
1208109712.203740 >> iscsid[3087]: child 3088 exited with status 0
1208109712.203781 >> iscsid[3087]: discovery process 0x9a070c8 pid 3088 fd 7 exited normally, marking for removal
1208109712.203839 >> iscsid[3087]: done starting discovery processes
1208109712.203887 >> iscsid[3087]: poll hangup on index 0, fd 7, discovery 0x9a070c8
1208109772.239283 >> iscsid[3087]: removing discovery 0x9a070c8, pipe has closed
1208109772.239294 >> iscsid[3087]: stopping discovery process 0x9a070c8
Seems like the md3000i is closing the connection, but im running out of ideas.
What is name of your iscsi initiator?
I can see iqn-
the - is not correct at that place.
The iqn must have the right format (with dot i.e iqn.2000-05.xxx:xxx check there:
http://wikis.sun.com/display/StorageDev/IQNNameFormat )
Your right. My iscsi initiator diden't follow the naming standards.
Works like a charm now.
Thanks for your time and help!
Glad to help. Sometimes the small things make the difference
Are you still getting the LegacyMP Plugin Errors?
I am having a similar problem with my MD3000i... how can I set my iSCSI adapter names back to the correct standard? iqn-xxxx etc.
I've tried disabling the iscsi adapter both via the GUI and console.
I have another server which is working fine connected to the MD3000i so its something with this particular install.