Start a Conversation

Unsolved

This post is more than 5 years old

K

1599

December 26th, 2013 14:00

multimapped LUN to ESX hosts from Clariion Array

Hi

I assigned luns to a ESX  farm but i got this trouble, i have a clariion CX480-C and HP Blade server

ESX 1 has assigned 45 LUN's, ESX2  has assigned the exact same LUN's ,ESX 3 has assigned the exact same LUN's, ESX 4 has assigned the exact same LUNs, ESX 5 has assigned the exact same LUNs, ESX 6 has assigned the exact same LUN, i mean, all of them have the same luns assigned from the clariion array,  but  the ESX1 shows 45 luns (OK), the ESX2 only shows 43 luns (Not OK), the ESX3 only shows 44 luns(Not Ok), the ESX4 shows 45 luns (Ok)  ,the ESX5 only shows 44 luns (Not Ok) and the ESX6 only shows 43 luns (Not OK)

I dont understand and i a little bit confuse because i did the zoning,i configured one  storage group with the 6 ESX hosts and the 45 Luns, the vmware team told me that there is any issue with the clariion, i dont think so, but iam not sure.

is there any issue with the HP blade (drivers, firmware, parameters, etc)? or i made a mistake from the clariion side

I´ll appreciate your help on this


4K Posts

December 26th, 2013 17:00

"Rescan" didn't work?

December 27th, 2013 08:00

Hi

I rescaned from the ESX side many times, but it does not work, is there lun reservation from the clariion side? is there a drivers or firmware problem from the HP blade?

any idea?

1K Posts

December 27th, 2013 11:00

If you select ESX2, go to Configuration and Storage adapter. Select the vhba and sort by the LUN column. Which LUNs are missing? The LUNs that you see there will be the HLUs.

Try removing ESX2 from the storage group and perform a disk rescan. Does the vhba see any other LUNs? Put the host back into the storage group and perform a couple of disk rescans. Are the same LUNs missing again?

1 Rookie

 • 

20.4K Posts

December 27th, 2013 11:00

put it in maintenance mode first of course

1K Posts

December 27th, 2013 12:00

Yeah, good point. Forgot that "minor" detail. Definitely put it in maintenance mode

December 29th, 2013 00:00

A few things that might help if maybe (just a hunch) we are up against vSphere configuration maximums such as total number of LUNs and/or maximum total paths:

1) What version of ESXi (or if by chance a older: ESX) are the servers running?

If they aren't all running the same version please note them individually.

2) Before you presented LUNs from the CLARiiON, by chance were there already LUNs presented from separate storage?

If so, please note the number of existing LUNs (you've already noted the number from the CLARiiON).

3) How many paths are there from the ESXi to the storage array (or arrays)?

4) How large are the LUNs you are trying to present?

Relevance will be associated with the version of ESXi once you provide the information.

December 30th, 2013 07:00

this can help

i got this info from the Emcgrab

Host Name

temm4mty2e2bl01

Max Lun

256

Total Number of Local Disks

1

Total Number of Fiber Channel Disks

255

ESXi 5.1 (vSphere 5.1)

The max lun is 256 and the ESX has 255 + 1 local disks

any idea?

1 Rookie

 • 

20.4K Posts

December 30th, 2013 08:00

ohh..so you have 256 LUNs present in the storage group already ? That would do it, can you move VMs around (storage vmotion) to free up some LUNs, reclaim them and present new LUNs bigger capacity ?

December 30th, 2013 09:00

I have mix, 45 from clariion and the others from VMAX the total 256 Luns


December 30th, 2013 10:00

dynamox wrote:

256 max per host or per target ?

Per host.  For ESXi (with its early Linux foundation) has a total count for LUN and/or paths that can't be exceeded regardless of the number of arrays.

For example, the LUN count is controlled by the following Advanced (Software) Setting available via the following breadcrumb trail in the vSphere Client:

1) Configuration > Software > Advanced Settings

2) Disk > Disk.MaxLUN

By default it is 256 (and is also the total maximum).

However, if you exceed 1024 total paths (again per host regardless of the number of arrays) before reaching 256, than that would be the limiting factor instead.  So as an example for others that might stumble on this:

1) Lets say I have for each hosts I have (zoned) 8 paths to the VMAX and 8 paths to the VNX

2) Then I would not be able to reach 256 LUNs but instead could only present 128 LUNs

EDIT: Removed comment about vSphere v5.1 maximum paths per LUN (it is 8 when using iSCSI but used FC as an example)

December 30th, 2013 10:00

Perfect, thanks for the information.  That confirms my suspicion and would be why you are not seeing all of the LUNs.

Per v5.1 ESXi host, you have the following limits (which ever comes first):

1) Maximum LUNs (256)

2) Maximum paths (1024)

http://kb.vmware.com/kb/1003497

For ESXi 5.1, as seen in the section "Storage Maximums" you will see the following:

1) LUNs per Server: 256

2) Number of total paths on a server: 1024

1 Rookie

 • 

20.4K Posts

December 30th, 2013 10:00

i believe the limit is 256 per target, so that's probably not it.

December 30th, 2013 10:00

256 per server

1 Rookie

 • 

20.4K Posts

December 30th, 2013 10:00

256 max per host or per target ?

1 Rookie

 • 

20.4K Posts

December 30th, 2013 16:00

thanks, mixed up with an older RHEL Qlogic limit of 256 LUN per target.

No Events found!

Top