Rivan being down

[matzah]

Just curious how many other people are being effected by this "rivan" RAID failure/restore

[TahirMeer]

well, it sucks... i have only 2 customers on the server and both of them are very angry with another downtime. I am unable to do my freelance work since i am unable to access my site and customer data on the VPS...

I am getting a feeling that it is time to move on from JPC... every few months there is some problem is Rivan... And once again they (JPC) failed to notify me about the downtime...

[matzah]

Anyone else think this is ironic and related to the outage last Sunday that we were only given 2 hours notice about a "maintenance window"?

Also, can anyone explain how a RAID failure occurs? I understand RAID and how it works when 1 drive fails, but I can not imagine in my mind how a WHOLE RAID can fail at once and then take 27+ hours to restore

[saracen]

A whole RAID array can fail for any number of reasons, including power supply problems and controller failure. I've had that happen to my RAID 5 array.

But as for the restore times ... good question.

[EuroNut]

When you're sitting in your office or at home watching a new disk being formatted or Windoze defrag doing it's stuff on a drive that hasn't been defragged since goodness knows when, you still feel in control. You can check back on the screen and see how many percent it's done, and plan your day accordingly. Go have dinner, watch a movie, whatever. You decide how often to look. No-one else is bothered - it's just you and your PC.

I'm not trying to talk down to anyone, or antagonize you, but these servers are a whole different game.

When a disk in your PC eventually fails, you can't do much else but pull it out, throw it in the trash, buy a new one, and restore all your precious data from the backup you did last night (OK, I won't even ask).

Having to wait while The Jag Techs find out which disk in a raid array failed, swap the disk, run a FSCK, watch the FSCK run to 90% then fail, find the backup, tell that to restore and overwrite all the data on that big array of drives, etc, etc, etc, is a pain, but what do you want them to do? How about they plug a monitor into the particular server and put a webcam on line so we can all watch the "percentage done" drift from 1.1% to 1.2% to 1.3% ...........

Frustrating as it is guys, and however good Jag build their servers, from time to time things WILL go pear-shaped. Because they're currently upgrading kernels and doing various other upgrades on the servers, THIS is the time when the gremlins will strike.

Just as we all hate it when we go check our PC for emails one morning and find that it crashed during the night, we really can't expect every server in the 2 data-centers that Jag uses to stay running for perpetuity.

It doesn't matter who you host with, this situation WILL occur one day. At least JaguarPC are transparent and tell us when something goes wrong, and what they're doing about it.

I vote for cutting Jag, and especially the techs, some slack - They have enough on their hands, and they're responsible enough to know, and sympathize with, the problems this causes for customers.

Just my 2 cents in support of JPC, who seem to be getting berated more than usual at present.

[Ron]

A restore shouldn't take 20+++ hours

[matzah]

EuroNut

I have tried really hard to not slam Jag, as I am not trying to blame them, but I trying to decipher what is going on and why.

I have run RAID systems in the past (definitely not as large as this), but trying to figure out how 31+ hours of restore (from the moment of failure, not officially just restore time) seems like this RAID array is for a lot more customers than I ever imagined or more than one server... if not please feel free to correct me

[TahirMeer]

Fallowing is the reply from Robert to my query.

the restore of the main /vz partition which is usually above 200 gb occurs from 80% to 100% , thus its very slow during this part of restore.

It has been more than 24hrs since this reply. But good thing is the new status say only 5% of the restore is left to be done... meaning more less 5hr minimum... as each 1% restore is taking more less 1-2hours

[matzah]

Things look up now....

Now it is time to determine what was lost and start playing interference with the customer base

[threehundred]

I'm on Sirius, and it has been down ever since Friday afternoon. It's been 36+ hours and we are still at like 80%. Not to mention the fact that I noticed someone said their backup was from Oct. 26. So not only is the server down, and its taking 30 hours to restore it from backup, but the backup is very old in technology terms.

I need to start looking for a backup plan for this.

[saracen]

Well, the marketing forVPS's does say that the service includes :-

"FREE daily backups"

If I'd lost days of data (which fortunately I have not) I'd be asking what happened to the more recent daily backups).

[TahirMeer]

Well, the marketing forVPS's does say that the service includes :-

"FREE daily backups"

If I'd lost days of data (which fortunately I have not) I'd be asking what happened to the more recent daily backups).

i just checked, and you are right the sales package does say "FREE daily backups", i would like to know the same what have to latest backup or a backup which was taken 2 days before the Raid failure ?

[matzah]

Anyone know how we can bring this thread to the attention of management... especially on the "Free daily backups" issue?

[TahirMeer]

Anyone know how we can bring this thread to the attention of management... especially on the "Free daily backups" issue?

It is pointless... i had talk with Edward... according to him only reliable backup they had was from 26th OCT. The backups from 1st and 2nd Oct were incomplete... (He didn't mention anything about days from 27 to 31 October) so i suppose those backups were never taken or were incomplete)

where as sales told me they only take backup once a week(on weekends), when asked why "Sales say Free Daily Backup" he says they take weekly system backup and daily account backups...

I am very much disappointed.

[signslave]

i'm on iri.nocdirect.com it seems to be down too. for a few hours now.