Im getting tons of errors from a network adapter. thought I would remove it....

[GodTheHamster]

So Im inheriting this vSphere environment and its a mess. im trying to sort through all the issues and im comin up short on this one.

Login to iSCSI target
iqn.2001-05.com.equallogic:0-8a0906-470014a05-fac000e7a364c4d1-sqllogs on vmhba36
@ vmk4 failed. The iSCSI initiator could not establish a network connection to the target.
error
4/2/2013 2:46:21 PM
scepvmh2.Domain.com

These repeat every 15 seconds or so, two of em each time for two different hosts. the target changes each time tho. but nothing seems to be having any issues, just the log fills up really fast and crashes vcenter. So i decided to remove the NIC from the scsi adapter, and i get this

Call "IscsiManager.UnbindVnic" for object "iscsiManager-49" on vCenter Server "VC01.domain.com" failed.
Operation failed, diagnostics report: iScsiLibException: status(80000000): unknown error; Message= IMA_VMW_RemoveNic

So im kinda stuck. these hosts are in production and if i break something....well i dont what to break something.

any one seent his before? any thoughts?

Thanks,

Dustin.

[solstice]

Did you check that the iSCSI (Array/Host, didn't mention your setup!) is up and functioning?

It sounds like either the switch between or the array is down / unaccessible.

[GodTheHamster]

The guy who was here before me said the errors started after deleting an unused/unneeded lun.....obviously it was doing something cause now it cant connect to it. but im having trouble just removing the network adapter from the iscsi config.

[solstice]

Just to get to the bottom of this, what kind of SAN array you using? Dell (MD/Equalogic?), NetApp?

Then you can check on the configured servers under configuration->storage adapters and do a rescan to see which ones are available and which ones aren't.

You can just remove the "unused" LUN from there.

[GodTheHamster]

Im thinking it could be a routing issue. The two adapters that are giving me error are intel adapters connected to the same swtich. the other two which are broadcom are connected to a different switch and do not seem to have any problems.

The network im working on its kinda peiced togeather, it was never set up correctly and I dont have the knowledge to fix it in full confidence known i wont mess it up.....reverse engineering is sometimes very difficult.

So the iscsi adapter is tagged with vlan50 and it seems to be having trouble with all the storage devices. the error changes which target. I think im getting off track here.

vmhba## has two nics, both show compliant, but one is giving me connection errors to all the datastores. every 8 seconds or so.

Edit: OK so i think ive found the problem. I was checking the switches and saw two tagged with a vlan but shows no link, so i checked the cableing and sure enough they are the uplinks to the san, all lights out. only one nic to the san is working.

Oh and they are Dell Equalogics.

Now i have to figure out how to log into the thing. its vland on a different subnet......blah.

[GodTheHamster]

well so far im just running in circles. I still dont know whats causing this error. found a few articles pertaining to a bug (thats been fixed) which would happen when one atapter would contastantly try and fail over. this seems simmilar but i dont think it is, as im never loosing connectivity to my SAN. just one of the atapters contstantly trying to log into the datastores.

One control cluster on the san is dark, in standby mode, the (v)NICs on the hosts show they are active but no link lights on the (p)switches. Ive tried setting the nics in the port groups to inactive, ive tried chaging ack time outs.

So vSwitch properties group 1

And hers iscsi2

here is a page that seems to explain the issue im seeing, but its not exactly....

http://vmtoday.com/2012/02/vsphere-5-networking-bug-affects-software-iscsi/

i duno, at this point im stumped.

Anyone have any insight, or questions that i can try and clarify?

[twilkinson2010]

This is an old thread, but since I just ran into this same issue and could not find any reference other than this, I'd like to offer my findings in hopes it will help someone else.

While cleaning up our environment, I found a vSwitch that was not being used - or so I thought. I deleted it and immediately started getting this exact error. It turns out that it was associated with the software iSCSI HBA, even though there were no devices connected to it. Unless the vSwitch has been removed from the Port Bindings (iSCSI Software Adapter > Properties > Network Configuration) it will create this error when you delete the vSwitch. I recreated the vSwitch and then properly removed it from the Port Bindings and then deleted the vSwitch. All better...