EW56 node temporarily down.

[jennifer]

Dear Customers,

The RAID HDD is degraded on one of our Windows node - EW56. We are working on replacing the HDD and resyncing the RAID array.

We will update this thread throughout the process. Your patience and co-operation is appreciated.

[Jack]

Hello,

We have replaced the Hard Drive on EW56. We are in process resyncing the RAID array however this will take some time as there is issue with one of our switch which will be replaced shortly.

We will keep you posted with the progress. Thank you all for your co-operation.

[Scott]

Hello,

The HDD has been swapped and the sync of data completed now. Everything is working fine on EW56 now.

[sihost]

Hi, question based on raid

I'm going to be buying a dedicated server with Raid 1. I noticed on this thread it has raid and was temporarily down.

When using Raid and there is 2 drives mirrored with the same information, I thought that if one fails then the server stays up while the failed drive is replaced, then it syncs the two again.

All this time customers would notice no downtime? Could you fill me in on this as if I'm wrong in thinking this then I probably don't need to pay the extra to buy raid because the reason I was going to do it is prevent downtime if a drive was to fail.

I noticed on a couple of other threads the same too when raids didn't prevent downtime.

Advice appreciated

[Jack]

Hi, question based on raid

I'm going to be buying a dedicated server with Raid 1. I noticed on this thread it has raid and was temporarily down.

When using Raid and there is 2 drives mirrored with the same information, I thought that if one fails then the server stays up while the failed drive is replaced, then it syncs the two again.

All this time customers would notice no downtime? Could you fill me in on this as if I'm wrong in thinking this then I probably don't need to pay the extra to buy raid because the reason I was going to do it is prevent downtime if a drive was to fail.

I noticed on a couple of other threads the same too when raids didn't prevent downtime.

Advice appreciated

You may find detailed information over here :- http://www.eukhost.com/forums/f15/raid-explained-4063/

Hope this post by Rock helps you

[sihost]

Thanks Jack, interesting read, that confirms my original thoughts on raid, in that it avoids downtime or data loss when a drive fails.

Why did this node go down for a couple of hours then while the drive was replaced and the raids resynced?

I thought the drive would be replaced and the data resynced and the only downtime would be restarting the server after the drive is replaced (like it says in rocks info)

Cheers

[Scott]

Hi,

We have taken all VPS offline to perform offsite backup before proceeding with HDD swapping. Server was offline because the switch on the same rack had problems and it is replaced by network engineers.

[martynh99]

Hi,

We have taken all VPS offline to perform offsite backup before proceeding with HDD swapping. Server was offline because the switch on the same rack had problems and it is replaced by network engineers.

Whilst i'm not on this node, this raises a few questions for me

1) In the hardware features for Windows VPS you state 'NAS Backup Storage'. A NAS device is 'offsite' to the server as it is a network device so why would the VPS have to be taken offline to facilitate this backup. Wouldn't you just need to do the normal weekly container backups then ship these across the LAN to the NAS shares ?

2) Server was offline because the switch was broken - don't the servers have dual network connections to seperate switches in a 'teamed' configuration - thats basic stuff for servers and has been for years.

[sihost]

speaking of which, have dedicated servers got this? Thinking of the 100% power and network uptime guarantee?

[John]

Whilst i'm not on this node, this raises a few questions for me

1) In the hardware features for Windows VPS you state 'NAS Backup Storage'. A NAS device is 'offsite' to the server as it is a network device so why would the VPS have to be taken offline to facilitate this backup. Wouldn't you just need to do the normal weekly container backups then ship these across the LAN to the NAS shares ?

2) Server was offline because the switch was broken - don't the servers have dual network connections to seperate switches in a 'teamed' configuration - thats basic stuff for servers and has been for years.

Hi,

NAS backup storage is where the backup is stored. When a RAID array fails, it takes a certain amount of resources to rebuild it. All VPS running and a raid array building can put a strain on the remaining disk drive. The last thing we want to happen is the remaining hard disk to fail during a resync. Therefore, if there is a high disk load, VPS are taken offline to allow the raid array to sync up again quickly.

Switches failing are very rare. You are welcome to read through the forums and find other examples, but I don't think you will find more than 5 in over three years. When you consider the number of switches we have, that is a good number.