How to replace physical disks

[p-pasquali]

Hello everyone! Sorry for the grammar, I use the Google translator.
I have a failing physical disk problem on an HP Proliant DL380G6 server. The server has:
- 2 SAS 146gb disks
- 5 SAS 300gb disks
there is ESX 4.0.0.
For some time now 1 146gb disk and 2 300gb disks have been blinking with the orange light, in addition to the green light. I assume they are slightly damaged, but I cannot find from vSphere Client where to see the status of the physical disks. I can see the cpu, the ram, the temperatures, power supplies etc .. but I can't see the status of the disks, and I can't even find where to read and manage the Raid.
Could you tell me on vSphere (or other) where I can see and manage the physical disks? In any case, could you tell me how I can replace them? Can I hot swap them? Do I need to unmount the VMFS first? Or do I really have to shut down the server? Do I need to replace one at a time? In this case, where can I monitor the rebuild of the raid?
Thank you very much for the valuable information !!!

[E_Espinel]

Hello.
With the following command we verify the version and the distribution installed can be VMware or HP

#esxcli software profile get

(you must enter in a remote SSH console that allows to enter commands)

If it is HP it is very likely that you have installed the utility to manage the disk controller. HPE Storage Controller Tool, this name has changed over time.
check if any of the following directories exist
opt/hp
opt/hp//hpssacli
opt/hp//hpacucli

If you found the last directory try running
# /opt/hp/hp/hpacucli/bin/hpacucli

=> ctrl all show config

 

[a_p_]

Wow, the classic ESX 4.0. I haven't seen this in a while. I wonder what the hosts uptime is?

I assume that the two 146GB HDDs are RAID1, and the five 300GB HDDs are RAID5.

What you may do, is to connect to the host's iLO Management, which should show you what's wrong with the HDDs.

If the RAID sets are managed by a P4x0 Smart Array controller, you can hot swap the disks (one at a time until fully restriped for the RAID set with two critical disks). Especially for the RAID1 you should actually replace the disk online, so that the controller doesn't accidentally mirror the new (empty) disk to the healthy disk.

Since the RAID restriping - after replacing a HDD - puts a high load on the disks, it's critical that you have backups of the VMs. It could happen that the second blinking disk (on the RAID5) faisl while the first one is restriping.

One more hint: After pulling out a HDD, wait a minute before you insert a replacement HDD, so that the controller can safely detect the missing disk!

André

[p-pasquali]

Thanks a lot for the replies! I apologize if I answer late ..
I reply to e_espinel:
i am not an SSH expert, however i searched for the main commands, and i tried to look for the folders you told me opt / hp. It will sound strange, but it seems to me that there are no subfolders in root, as you can see from the attachment! Could be? Or am I wrong to use SSH commands?

I answer to a_p_:

Yes, it is a server installed in 2008!
If I'm not mistaken an HP server's iLO should have an independent physical network connection, right? If so, I know that server has only one network connection (only one lan cable). So I assume it doesn't have iLO!

Do you think I can do something to see the status of hard drives, and raids? Or do you think it can proceed to replace "hot" the individual hard disks? I can also physically go to the server, it has a keyboard and a monitor.

Thank you very much!!!

[E_Espinel]

Hello.
On the ProLiant DL380 Generation 6 (G6) server, you have a separate iLO remote management port (RJ45), it is located on the left side of the serial connector (DB-9) should have a mark ILO.

You should try to connect a network cable to the ILO port and try to access with  the factory defaults, but if you don't succeed you will have to shut down the computer normally and then enter in the Bios to configure the ILO (password, IP).To enter the BIOS of an HP server usually during Start press F9.

ILO Factory defaults
By default, your iLO interface is configured to use a DHCP IP address.
On the DHCP server, you need to look for a device using the name listed on the HP TAG.
From an older browser (preferably 32bit) try to enter the ILO IP address
On the prompt screen, enter the administrative login information.
Factory default access information:
Username: Administrator
password: the password listed on the HP tag

HP tag sample

each HP server has its own tag with its different and unique data, you should look for the HP tag on your server.

 

[niyijr]

 The DL 380 G6 should have ILO2. I remember working on the DL 380 G5 running ESX 3.5 years ago.

As you follow  steps, check out this video. I've found that a video walkthrough helps.

Good luck!

[p-pasquali]

Thanks so much! I managed to enter iLO, it was configured in DHCP, so I just needed to find the ip and enter with the password contained in the tag.
It actually shows me 3 drives in fault! I am attaching an image. But I can't find any menu from iLO where I can monitor or edit the raid. Is it possible that the raid can only verify this from the server console, or from the remote console? Without the license I can't use the remote console, can I?
Thanks again everyone!!!

[E_Espinel]

Hello.
As far as we can see in the ILO, we have 7 disks in the disk bays (slot 1 to slot 7)
marked with failure are disk bays 1, 4 and 6.

If we assume that the RAID scheme was
2 disks of 146GB  (type EG0146FAWHU ) in RAID 1
5 disks of 300GB (type EG0300FAWHV ) in RAID 5

The first disk to change would be the 146GB disk which should be in disk bay 1,
The fault LED should be on or blinking (amber), the activity/line LED should be on or off (green).
If both LEDs are blinking, the disk cannot be removed, some process is in progress.

When removing the disk you must wait at least 1 minute for the controller to register this change. Then you must insert the new disk and wait for the rebuilding process to finish, this may take a few hours. At the end, only the activity LED (green) should be active.

In RAID 5 you would have 2 failed disks (this is very serious), from what you can see in the ILO, you would have 4 disks for Data and one disk as Hot spare. So in summary 2 disks failed so the hot spare is being used.

The hot spare is a spare disk that is not being used, and when a disk in the array fails it is automatically activated and comes into use.

The failed disks would be in disk bays 4 and 6.
Proceed as above, but only one change at a time, it is recommended to wait 10 to 15 minutes between changes.
During the rebuilding process the activity (LEDs) of the disks in that array will have more activity than usual.

You can monitor the status of the disks from the ILO, when the changes are finished the status should change.

You will need to order new disks from your hardware vendor, using the spare part number on the front label of each disk.
Check the failed disks in disk bays 1, 4 and 6 directly.

Based on what you see on the ILO they would be
one HP Product ID EG0146FAWHU 146-GB 6G 10K 2.5 SFF SAS disk.
two HP Product ID EG0300FAWHV 300GB 10k 6G 2.5 SFF SAS disks.

All disk work needs a full backup beforehand of the data or take it to an external storage.

The disk replacement is not complicated and is online, but if you are not experienced in this, you should hire an HP service provider.

 

[Dennis898]

Select the disk group that contains the device that you want to replace. Select the flash capacity device or magnetic disk, and click Remove selected disk from disk group.

myaarpmedicare