Surprise found on server

[Dujith]

I am in the process of updating some software on the server we need for ordering components and such.

While doing that i noticed that our disk space is getting low. So i check a few things and was curious on what was in the actual server hardware wise.

 

So fired up the remote desktop logged into the main server from where the other servers run in hyperv and look at what ssd's are installed.

I see 4x 149GB SSD (which was a normal number when this thing was build) but then i noticed the complete size: 595GB....

 

A quick check confirmed, the IT company that built our server has 4 SSD's in raid 0..... no wonder it took em a long while to rebuild after we had a drive failure in the past.

 

I hope this is not normal an i have something to shout at them this monday.

 

On a side note: This IT company has been annoying us for a while now for various reason, so this would be a good point to swap companies

We are a small company with most ppl on the road and just 3 persons in the office working on computers. How hard would it be to backup the VM's make a new raid from newer SSD's put em in Raid 5 i suppose and put back the VM's. I guess not hard at all since the main machine has its own SSD for Server 2012 and runs the Servers we work of in HyperV.

[Jarsky]

There should be nothing to 'rebuild' when its a RAID0 as theres no parity. Are you sure you aren't looking at the physical layer? Generally on servers you'll be presented the physical storage which will show you raw stats, and then the logical which will have the RAID volume. 

 

What model server is it? 

[_Dr_Eye_]

2 hours ago, Dujith said:

I am in the process of updating some software on the server we need for ordering components and such.

While doing that i noticed that our disk space is getting low. So i check a few things and was curious on what was in the actual server hardware wise.

 

So fired up the remote desktop logged into the main server from where the other servers run in hyperv and look at what ssd's are installed.

I see 4x 149GB SSD (which was a normal number when this thing was build) but then i noticed the complete size: 595GB....

 

A quick check confirmed, the IT company that built our server has 4 SSD's in raid 0..... no wonder it took em a long while to rebuild after we had a drive failure in the past.

 

I hope this is not normal an i have something to shout at them this monday.

 

On a side note: This IT company has been annoying us for a while now for various reason, so this would be a good point to swap companies

We are a small company with most ppl on the road and just 3 persons in the office working on computers. How hard would it be to backup the VM's make a new raid from newer SSD's put em in Raid 5 i suppose and put back the VM's. I guess not hard at all since the main machine has its own SSD for Server 2016 and runs the Servers we work of in HyperV.

that would not be very hard or time consuming but why, should use virtualbox or vmware

[Dujith]

1 minute ago, Jarsky said:

There should be nothing to 'rebuild' when its a RAID0 as theres no parity. Are you sure you aren't looking at the physical layer? Generally on servers you'll be presented the physical storage which will show you raw stats, and then the logical which will have the RAID volume. 

 

What model server is it? 

With rebuild i meant the actual time they were here working on it. Not the rebuild raid thingy  sorry if i got the terms mixed up.

I attached the screenshot from the adaptec manager.

1 minute ago, _Dr_Eye_ said:

that would not be very hard or time consuming but why the fuck r u useing hyper v u should use virtualbox or vmware

Cuz its installed by that company? Also its not like we are the only ones running HyperV  That however is not part of this discussion so please dont turn this into a HyperV sucks thread  

[Jarsky]

Ah yup, definately a RAID0. 

 

Is it just the VHD's that are on the RAID? If so, just power them down, copy/move the VHD's to a backup disk. 

Destroy and rebuild the RAID as a RAID5, then move them back. 

 

You could edit the virtual disk location in the VM and use a backup disk to start the servers back up and run them temporarily until you've finished rebuilding the raid. 

Though destroying and re-creating a RAID5 is really fast - so could easily do it on a weekend afternoon/evening.  

 

Edit: If you have snapshots, or VM settings etc...then use the 'export' option to copy it to the other drive.

[Dujith]

@Jarsky We have no snapshots as Veam creates a daily, weekly and monthly backup to a seperate NAS (which is RAID5  )

I'll copy the VHD's to there and i think i'll just remove the 149GB SSD (Intel 320) and replace em for 4x Intel 540, 480GB

Which should net me 1,4TB.

 

Any other SSD to suggest? Workload would be mainly documents and 1 program that runs of the server for everyone. 

Or is the Intel a good choice for us?

[leadeater]

@Dujith

 

You'll be able to just add 4 new SSD to the server, create a RAID 5, Live Mirage the VMs to the new storage, delete old volume/partition, delete old array and then remove the old SSDs.

 

You can do all of this with no down time, assuming free hotswap bays in server.

https://blogs.technet.microsoft.com/canitpro/2014/04/23/step-by-step-completing-storage-live-migration-in-hyper-v-2012-r2/

 

For SSDs Intel or Samsung are the ones I recommend.

[leadeater]

31 minutes ago, _Dr_Eye_ said:

that would not be very hard or time consuming but why the fuck r u useing hyper v u should use virtualbox or vmware

Hyper-V is fine and actually very good, Virtualbox shouldn't be used for server hosting for production servers.

[Dujith]

@leadeater "ass"?  i'll need lube for 4 SSD's, 1 i can manage... but 4.....

[leadeater]

1 minute ago, Dujith said:

@leadeater "ass"?  i'll need lube for 4 SSD's, 1 i can manage... but 4.....

lol add, ass is better tho 

 

Also if you don't use Live Migrate you'll mess up the Veeam backups as they'll be 'new VMs' and the backup chain will be broken causing both a new full but also making it a pain if you need to do a restore and it's from around the time do did the changes.

[Dujith]

@leadeater Ah, i would have never thought about that. I'll first shout at the IT company on monday then either they will fix it or i will, keeping this in mind.

[Dujith]

@Jarsky @leadeater

Small update:

IT company is gone since they had:

a. no explanation as to why it was running raid 0.

b. pricing: more then twice the amount for the ssd's i could find at a normal retailer. And for some reason it would take em a week. 

So bye bye.

 

The Export like you said forced itself on me anyway as room was out and the backup started to fail.

So i did what u said: Export to NAS, Delete VM on ssd, Import at NAS location so it could finish the merge. Then in the Hyper V manager i just used the move wizard and now its back and running on the ssd.

 

I ran the Veeam backup job after this and it just made an incremental backup so thats working fine too

 

In light of this i will just do the ssd upgrade myself as it was very easy and just took time.

 

Now just have to check which ssd's to use and i'm good to go.