This post is more than 5 years old
13 Posts
0
202350
July 10th, 2013 04:00
Path redundancy warnings in VMware
Hi
We have noticed recurring warning events in VMware similar to this:
Lost path redundancy to storage device
naa.6090a0b8d08e35c1bd16d5ded001507f. Path
vmhba41:C1:T24:L0 is down. Affected datastores
: eql_vmfs-vdi-thinapp.
warning
10.07.2013 12:18:48
eql_vmfs-vdi-thinapp
Followed by:
Path redundancy to storage device
naa.6090a0b8d08e35c1bd16d5ded001507f
(Datastores: eql_vmfs-vdi-thinapp) restored. Path
vmhba41:C1:T24:L0 is active again.
info
10.07.2013 12:19:00
eql_vmfs-vdi-thinapp
And also:
Alarm 'Cannot connect to storage' on esx04.hials.
no changed from Gray to Gray
info
10.07.2013 12:22:51
esx04.hials.no
Firmware EQL: 6.0.2
Esxi version: 5.1.0-799733
Software iscsi adapter
We have installed MEM and also disabled delayedack, LRO and increased login_timeout to 60s.
Any tips?
Origin3k
6 Operator
•
2.4K Posts
•
12.3K Points
1
July 10th, 2013 05:00
Welcome to the club,
we seen this on every vSphere+EQL solution we deployed over the years. Most customer dont see it because the havent configurated the special Alarm notification on vCenter.
The "problem" occurs once a week or month. Most of the time not all of the datastores are effected and also not every Host. The reconnect is done in the same second and we never see greater problem there. The bad this is that you get the false-positive alarm mails.
Regards,
Joerg
hkraal
1 Rookie
•
41 Posts
1
July 10th, 2013 05:00
This is normal for storage systems which manage the network load across their ports as EqualLogic does:
kb.vmware.com/.../search.do
However I would certainly look into this further since it happens every 5 minutes which would make me suspicious as this could be the result of an other problem like wrong settings or perhaps a congested network?
standal
13 Posts
0
July 10th, 2013 05:00
Thanks for the info Joerg.
I'm seeing this warning much more often, one warning every 5 minutes or so..
Added some new datastores this week, might this be the reason for the increased volume in warnings?
We have also installed Veeam System Center MPs for VMware recently, so this might be the reason for why we became aware of the warnings... :)
Regards,
Standal
standal
13 Posts
0
July 10th, 2013 06:00
We are running 6.0.2, so an upgrade to 6.0.5 might solve this
Awaiting the what the support team finds out before upgrading though.
Thanks Don.
Regards,
Standal
standal
13 Posts
0
July 10th, 2013 06:00
Thanks hkraal,
I've just opened a case with Dell Eql support
I will post the results in this thread.
BR
Standal
standal
13 Posts
0
July 12th, 2013 05:00
Dell support rep. also recommended to upgrade to firmware 6.0.5
Will Schedule this in next maintenance window.
BR
Standal
tarquel
5 Posts
0
July 23rd, 2013 04:00
We also get this sort of error logged on our setup too and its something i've been trying to get my head around off and on for some time so here's where I'm at with it.....
To try and stem a new issue this past weekend with a large VMFS5 LUN where a large VM was vmotioned onto this lun and was then having serious freezing/lagging issues, I thought it was prudent to upgrade to the latest firmware from EQL FW 5.2.2 and did so last night to the latest version (currently 6.0.5).
Happily, the VM is working much better but alas, it made no difference to the errors in the host logs...every 5mins or so we get it appear on at least one of the datastores on each host from what i can see. Its not always constant however... i can see periods of an hour or so where there is no errors of this nature.
For info on the setup, we have 3x PS6500's in the same cluster / storage pool and (currently) 3x 5.1 vsphere hosts (just shy of the latest set of updates) that are PowerEdge R720's.
We have another DC with the same EQL arrangement but the farm is running R810's and is currently on 4.1 (to be replaced with 720's and vsphere 5.1 etc).
A while back (6 months or more) i did look to updating the drivers and firmware on the nics, to see if that helped at all - but it didnt, but possibly thats worth another look, as maybe theres some bug in the nic's firmware / vmware drivers perhaps causing more issues with the iscsi connectivity..?
Not trying to hijack the thread with my own woes though... just wanted to add my 2pence / 2cents :D
Cheers.
tarquel
5 Posts
0
July 23rd, 2013 04:00
Whoops, forgot to add in there that on the older farm (with the R810's and vsphere 4.1), we dont get the continual churn of errors.... EQL's there are running EQL FW 6.0.4, but i dont recall the errors messages on the hosts when they were running EQL FW 5.2.1, so I dont know in what factors the difference is....