HOW TO test HP ProLiant MicroServer Gen8 RAID 1

[arcticeye]

Good moring,

I have my HP server installed with ESXi 5.5 U1, and also the raid 1 array already configure, and supposed to work. The System Base installed in a micro SD card and the hdd array only for the vm's.

I wonder is there is any way to test how the RAID 1 works. IE: Can I take one of the disks out and i will find the information on the other intact? What If i do it while the system is connected? If i turn off the server, take out 1 disk, then turn on, create a file "disk2" therefore i turn of and put again the disk and take out the other and create a file "disk1", when i put the 2 disks should I have the 2 files??...

I'm giving examples and I would like if you could explain a little bit about this...to be sure when the RAID wil work and when NOT.

Thank you very much,

Kind Regards,

Matias

[vervoortjurgen]

hello

well only test i always do is to unpkug one drive and see if everything stays up and running

youll need raid controller card or raid software to see content of the other drive on other pc

or if you have the same exact server you can read the config from disk and see if your datastore is there

i turn off the server, take out 1 disk, then turn on, create a file "disk2" therefore i turn of and put again the disk and take out the other and create a file "disk1", when i put the 2 disks should I have the 2 files??..

-> doesnt work that way. one disk will be over writen

-> RAID 0, RAID 1, RAID 5, RAID 10 Explained with Diagrams

i rarely test it like that. its so commenly used that you accept it that it would work.  never had any failure or problems that its not working.

[arcticeye]

Hello Vervoort,

Thank you for your answer.

So, in conclusion, I just take out a Disk (no matter which) from the array and it should continue working.

If one of the disks breaks while the system is working, it's supposed to continue working and advice me about the broken disk. There are now 2 questions:

1) Isn't there a change that the broken disk data be copied to the healthy one and then the FS of the healthy one be corrupted? i imagine there's a protection that will stop this kind of attempt. Is like that?

2) If one disk breaks I have to turn the server off and then add a replacement disk (when I get it) and it will work all alone?? Or I will have to enter to the raid bios console to reconfigure the new disk to the raid 1 array?? Is there any procedure to do that in my HP proliant Microserver G8?

Thank you again,

Kind regards.

Matias

[vervoortjurgen]

yes the raid controller can send notifications. most branded products show a warning light on the disk

1) no the raid controller marks the disk as failed and shouldn't replicate from bad disk to good disk

2) read your specs if you raid controller supports hot plug drives.  all branded servers (HP, IBM, cisco,...) support almost hot plug drive. meaning you can change the drive while its running.  if you replace the disk then normally it would rebuild your raid config. but always double check it in the raid controller console.  or you can set it up as spare first.

[arcticeye]

Dear Vervoort,

Thank again!I

One more thing though, suppouse a disk fails and I take this out and insert a "healthy" one but that's got a lot of data inside. Does RAID deletes all the data and prepares it automatically in most Brand new systems as the HP Proliant microserver G8?

Kind regards,

Matias

[cykVM]

That depends on the controller config, if it uses newly added devices/disks for automatic rebuild or if they are put in as unknown or spare first.

If the rebuilt is started the data on the replacement drive will get overwritten.

But one would not use a drive with (important) data as a replacement anyway - at least I won't do that.

[arcticeye]

Hi CykVM!

Thanks for your answer.

I tried to replace the second disk by another and I get this message. The new disk isn't a 2TB as it says, it's a 1TB disk. It's supposed to be a healthy disk.

What do u interpret? It's well plugged in so I don't think this will be the issue.

Thanks

Kind regards

Matias

[cykVM]

Sorry, I don't fully get what you say.

Is the replacement drive a 1TB or a 2TB?

What size is the remaining drive. Usually you replace drives only with same type/model.

[arcticeye]

Dear CykVM,

The disk was broken, my mistake!

I will get a Brand new disk with the same Brand/space than the ones I have and then test it.

Thank you for your help again!

Kind regards,

Matias