Problems Creating an RDM

[pviqueira]

Hello,

I have this scenario:

- 1 ESX Server 3.5 Upgrade 1 in HP DL 380 G5 connected to a SAN HP MSA500 G2

- Virtual center 2.5

I have two storages in the SAN. The picture below show this storages:

I present another LUN in NTFS format to ESX Server and I want a VM see this LUN as a RDM.

If I make a rescan hbas in ESX server. I can see that the new LUN in NTFS format is presents to ESX server.

I go to the VM to add a new disk (Raw device mapping) but I see this option disable.

Any Idea?.

Yhanks

Pablo

[azn2kew]

What steps did you create your RDM SAN LUNs? Did you correctly present it to ESX cluster and make sure the VM you're adding disk can see that RDM as well. What happen if you make VMFS partition? NTFS should works under ESX because its RAW data pass through.

If you found this information useful, please consider awarding points for "Correct" or "Helpful". Thanks!!!

Regards,

Stefan Nguyen

iGeek Systems Inc.

VMware, Citrix, Microsoft Consultant

[kjb007]

Verify what LUN ID you gave that disk. Check your vmkwarning and vmkernel files. ESX may have disabled access to the LUN because of resignature. Check for the string resignature or snapshot in the logfiles.

-KjB

[pviqueira]

Thanks.

The problem is that the ESX sever "see" the LUN as we can see in second picture, but when I want to present this LUN as a RDM to my VM, this option is disable (third picture).

Steps:

1. With HP ACU (Array Configuation Utility, software to manage LUN in MSA500) I present a LUN in NTFS format (this LUN contain datas that I can see from a new VM) to the ESX server.

2. I rescan hbas an I can see the new LUN presented . (Picture 2).

In picture 2, target 1 and target 2 are vmfs storages and target 0 is the LUN in NTFS format that I need to present to a VM (this LUN contain datas from a physical machine).

3. I try to present this LUN to a virtual machine adding a new disk and selecting RDM, but this option is disabled (Picture 3).

Do I need to do something else?.

Regards, Pablo

[kjb007]

The thing is, that ESX will only allow LUNs it finds are available to be added as either storage for itself or as an RDM. If it does not see it "available", it will not allow you to add it. In other words, as I mentioned, even if ESX sees the LUN, if it believes the LUN is a snapshot, or from another host, or has the same LUN ID as an existing LUN, it will mark it as a snapshot, and disallow access to it. If it does that, it will add warnings in the vmkernel and/or vmkwarning logfiles to let you know that it has disabled access to that LUN.

So, even if it can see the LUN, it may have disabled access to that LUN, making it unavailable to be added as storage or an RDM.

Hope that makes sense,

-KjB

[pviqueira]

But, the LUN is available to add as a storage:

But I don´t know the way to make it avalaible as RDM.

Thanks

[kjb007]

You said the LUN is formatted NTFS. Is it a new LUN, or are you moving it from another server, meaning is there data on it?

Do a couple of things:

1. Rescan your hba a couple of times

2. If you still can't add, run the below and post the output:

esxcfg-mpath -l

esxcfg-vmhbadevs

fdisk -l

-KjB

Message was edited by: kjb007: added additional steps

[mike_laspina]

Hi,

This occurred for me once, it's a fair while ago so I'm not sure exactly what the process was to correct it.

I recall that I had to remove the mapping from the hosts scan them all again so the LUN was not available as storage for the host.

The map it back on the host but do not scan for new vmfs volumes or storage

Then I could see the disk as RDM available in the VM add disk setting.

[pviqueira]

Thaks again.

I´m trying to to move the LUN from another server (the Lun was present to a physical machine with data, this LUN contain two NTFS partitions).

I rescan the hba again

and I try to add the RDS but this option is disable:

-

---

root@aldesx root# esxcfg-mpath -l

Disk vmhba0:0:0 /dev/cciss/c0d0 (69973MB) has 1 paths and policy of Fixed

Local 6:0.0 vmhba0:0:0 On active preferred

Disk vmhba1:0:0 /dev/cciss/c1d0 (295104MB) has 1 paths and policy of Fixed Local 18:1.0 vmhba1:0:0 On active preferred

Disk vmhba1:1:0 /dev/cciss/c1d1 (286101MB) has 1 paths and policy of Fixed

Local 18:1.0 vmhba1:1:0 On active preferred

Disk vmhba1:2:0 /dev/cciss/c1d2 (286101MB) has 1 paths and policy of Fixed

Local 18:1.0 vmhba1:2:0 On active preferred

-

---

root@aldesx root# esxcfg-vmhbadevs

vmhba0:0:0 /dev/cciss/c0d0

vmhba1:0:0 /dev/cciss/c1d0

vmhba1:1:0 /dev/cciss/c1d1

vmhba1:2:0 /dev/cciss/c1d2

-

---

root@aldesx root# fdisk -l

Disk /dev/cciss/c1d0: 309.4 GB, 309439182848 bytes

255 heads, 63 sectors/track, 37620 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

/dev/cciss/c1d0p1 * 1 23498 188747653+ 7 HPFS/NTFS

/dev/cciss/c1d0p2 23499 37620 113434965 7 HPFS/NTFS

Disk /dev/cciss/c1d1: 299.9 GB, 299999690240 bytes

255 heads, 63 sectors/track, 36472 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

/dev/cciss/c1d1p1 * 1 36472 292961276 fb Unknown

Disk /dev/cciss/c1d2: 299.9 GB, 299999180800 bytes

255 heads, 63 sectors/track, 36472 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

/dev/cciss/c1d2p1 1 36472 292961276 fb Unknown

omitting empty partition (5)

Disk /dev/cciss/c0d0: 73.3 GB, 73372631040 bytes

255 heads, 32 sectors/track, 17562 cylinders

Units = cylinders of 8160 * 512 = 4177920 bytes

Device Boot Start End Blocks Id System

/dev/cciss/c0d0p1 * 1 25 101984 83 Linux

/dev/cciss/c0d0p2 26 1053 4194240 82 Linux swap

/dev/cciss/c0d0p3 1054 3623 10485600 83 Linux

/dev/cciss/c0d0p4 3624 17562 56871120 f Win95 Ext'd (LBA)

/dev/cciss/c0d0p5 17037 17286 1019984 83 Linux

Regards, Pablo

[hillda01]

Hi,

We use HP MSA500 G2 with our ESX 3.0.1 hosts but dont make use of RDM.

I'm not sure RDM is available to use in such a set up as the HP MSA500 G2 is a SCSI based solution...

Do you have things like active passive HBA's working with SCSI?

Regards

Dave

[kjb007]

Have you also made sure that the former owner of the LUN no longer has access to it? Just to make sure there are no reservations still on the LUN?

-KjB

[pviqueira]

I have three servers acessing to the MSA (Two physical machines and one ESX server). LUN 2 was before presented tho a physical machine an now is present to ESX sever. I have only a path between ESX server and MSA500 G2 (MSA has 4 ports to connect server, 3 ports are busy)

I have presented the new LUN for RDM (unidad logica (LUN) 2) using the same way as other LUNs for vmfs storage (unidad logica (LUN) 6, unidad logica (LUN) 7).

The picture below shows the mapping between ESX server and MSA 500 G2.

Thanks.

Pablo

[kjb007]

I would try also what Mike.Laspina was suggesting above.

Let's clear things out, and restart. Edit the mask/presentation so no host has access to it. Rescan the hba's on your esx host and make sure the LUN is gone.

Then, update the mask/presentation to give access to the esx host again, and do as Mike stated.

Try and add the RDM and see if the option is not grayed out.

If not, rescan the HBA's, don't check the boxes. Refresh the storage section, and then try to re-add the RDM.

-KjB

[VCPGuru]

I had the same problem.....I had to wait approx. 20 minutes after the esx saw the RDM mapping

[ssrtist]

hey pviqueira,

please try this:

1. go to "configuration" tab of your esx server in VI Client

2. Advanced Settings

3. LVM

4. set LVM.Enable.Resignature to 1 (default 0)

let me know if this works,

Kenny (ssrtist = solid state artist)