Primary HDD failure on Concorde.eukhost.com - Page 2 - UK Web Hosting | Dedicated Server Windows and Linux VPS Forum

**eUKhost.com** · #21 (**permalink**) 01-12-2006, 09:15

If anyone has got any problems with their account other than post 18th November data should open a ticket through our helpdesk from http://support.eukhost.com and provide me with the ticket number over here.

I will deal with those tickets personally and get all your problems sorted.

**sprint66** · #22 (**permalink**) 01-12-2006, 09:42

Ok, so I don't know a huge amount about this, but why are such important drives not mirrored? Surely having a drive mirrored would have eliminated any major loss of data?

**ThePants999** · #23 (**permalink**) 01-12-2006, 09:57

Yes, I'm sure I saw you say on another thread that you ran RAID on Linux servers.

**ThePants999** · #24 (**permalink**) 01-12-2006, 09:59

Quote:

Originally Posted by eukhost.com

We cannot make daily backups on such servers as it will keep the server overloaded at least for 3 - 4 hours every night which will result in problems for server.

That's strange. My last host ran daily backups with no problem. Personally I think it's very important to do so.

Quote:

Originally Posted by eukhost.com

We started making multiple backups of all servers since then but now we need to make daily backups of mysql databases to avoid any such problem in future.

Could you possibly do SQL Server backups daily as well, on your Windows servers?

**eUKhost.com** · #25 (**permalink**) 01-12-2006, 10:09

Concorde is a 3 year old server and we never had RAID configured on servers at that time. all data from Concorde was getting mirrored using rsync on the secondary drive but something went wrong with crons which were set for backup or something else prevented mirroring on 26th.

We have configured daily database backup on all our servers last night but there seems nothing else that can be done to recover post 18th November backup.

We have other 3 old servers which don't have RAID configured and we will migrate them to new servers next week to avoid similar problem.

RAID is good in case of such drive failures but RAID has got some disadvantages as well. I don't think we should discuss those disadvantages right now as the priority right now is to sort database and email problems of those who suffered data loss due to HDD failure on Concorde.

**ThePants999** · #26 (**permalink**) 01-12-2006, 10:43

Quote:

Originally Posted by eukhost.com

We have configured daily database backup on all our servers last night

All databases on servers? Including MSSQL databases on Windows servers?

**eUKhost.com** · #27 (**permalink**) 01-12-2006, 11:35

Yes. All databases on windows servers as well as Linux servers.

**steveu6** · #28 (**permalink**) 01-12-2006, 11:42

Quote:

Originally Posted by eukhost.com

If anyone has got any problems with their account other than post 18th November data should open a ticket through our helpdesk from http://support.eukhost.com and provide me with the ticket number over here.

I will deal with those tickets personally and get all your problems sorted.

I'm trying to get in to trace the ticket number I was given last night, but last night I did it from my user name, but now from somewhere else your system wants an e-mail address.

My domain name is steveu dot org and the ticket was raised as I had lost everything. The web website and the e-mail addresses are now back but there is not trace of the mailing lists and mail to them is bouncing.

Unfortunately the e-mail confirmation of the raising of the ticket does not give a number. It would be useful if the confirmation mail had the details of the ticket on it.

http:// steveu.org/mailman/listinfo/dfhgc_steveu.org was the home page of one of my mailing list home pages.

**eUKhost.com** · #29 (**permalink**) 01-12-2006, 11:52

Hello Steve,

I am working on this and will restore your mailing lists.

Please allow some time for the lists to be restored.

**Fish_n_Things** · #30 (**permalink**) 01-12-2006, 12:08

Is it possible to get a cron job setup on the server to take a daily backup of 2 MySQL databases, zip (or Gzip) them up into a file with a date and time as it's name and place them in a folder so I can auto-FTP them to my own machine over-night?

**eUKhost.com** · #31 (**permalink**) 01-12-2006, 12:20

Yep.

We have similar setup for some other customers and crons have been setup for them. you need to open ticket for our senior system admins to do this as I don't have much knowledge of how it is done for other customers.

**eUKhost.com** · #32 (**permalink**) 01-12-2006, 12:24

Quote:

Originally Posted by steveu6

I'm trying to get in to trace the ticket number I was given last night, but last night I did it from my user name, but now from somewhere else your system wants an e-mail address.

My domain name is steveu dot org and the ticket was raised as I had lost everything. The web website and the e-mail addresses are now back but there is not trace of the mailing lists and mail to them is bouncing.

Unfortunately the e-mail confirmation of the raising of the ticket does not give a number. It would be useful if the confirmation mail had the details of the ticket on it.
http:// steveu.org/mailman/listinfo/dfhgc_steveu.org was the home page of one of my mailing list home pages.

Your mailing lists have been restored successfully.

Let me know if you get any errors with those as there can be permission problem which will need another cpanel update.

**ThePants999** · #33 (**permalink**) 01-12-2006, 14:44

Quote:

Originally Posted by eukhost.com

Yes. All databases on windows servers as well as Linux servers.

Thanks very much for that, it's appreciated - I'm much happier knowing the databases are getting daily backups. A week can be a lot of data to lose.

**steveu6** · #34 (**permalink**) 01-12-2006, 17:54

Quote:

Originally Posted by eukhost.com

Your mailing lists have been restored successfully.

Let me know if you get any errors with those as there can be permission problem which will need another cpanel update.

The mailing lists are back.

I will be checking things out and getting list membership up to date. But AFAICT 98% of everything is back.

Thank you very much for sorting this out.

I will have to learn to use backup via cpanel in future.

I know it will not have helped in this case unless it had been elsewhere.

Later.

**eUKhost.com** · #35 (**permalink**) 01-12-2006, 20:05

We understand how important backups are as such Hard drive problems have hampered our efforts to give top class service to our customers.

If we get at least some form of messages in logs or something on the monitoring then such problems can be avoided but this things keep on happening with servers and we have decided to give utmost priority to backup of databases.

In case of email service outlook should be given preference over webmail as you can keep a copy of emails in your outlook and emails are downloaded on your local machine but databases cannot be downloaded on daily basis and thats the reason we are giving priority to backups of databases.

We have started additional backup service to allow our corporate customers to keep a daily backup of their mission critical data. I don't think anything would be as good as this for those who are running dynamic websites.

If you wish to maintain a separate backup space on a remote server where you can connect through FTP and we can setup rsync mirroring then you should take a look at our backup service.

We will setup daily, weekly & Monthly rsync crons to copy backup of your accounts on your backup space. This service would be addon service and we wont compromise with backup service which is included in your hosting plan.

**g30rge** · #36 (**permalink**) 04-12-2006, 19:19

Would this hdd failure have affected anything in the mail programme?

Reason for asking is I have sent out various mails to groups of people and more than half of them are not receiving the mail.
I also have people telling me that mail they are sending to my account is being returned (sent last night)

**eUKhost.com** · #37 (**permalink**) 04-12-2006, 19:52

There should not be any such problem as none of our other customers have reported anything similar.

If you have a ticket open for this then do let me know the ticket number and I will have someone to look into it immediately.

**kanasta** · #38 (**permalink**) 22-01-2007, 02:00

why wouldn't you simply have RAID?
most other host's offer daily backups, and the cost of the hard disk is easily offset by the users who are going to need refunds in case of a failure

**eUKhost.com** · #39 (**permalink**) 22-01-2007, 11:27

Thats true.

We have setup all our new server with RAID 1 from year 2006 but in case of Windows Servers RAID is of no use.

We have RAID 1 on all Linux Servers and Daily, Weekly as well as separate monthly backups for all linux as well as windows servers. We have given high priority to backups from last year and none of our existing customers had any problems when they needed backup.

In case of concorde the primary drive failed completely in last week of November and we were shocked to see the daily and weekly backup crons failure for over a week. Customers lost 1 weeks data but it was not complete loss of data.

We started additional paid backup service after that to allow our corporate customers to have daily multiple backups of their data and they also get notified by the backup crons. If you need control over your backup then you should consider our backup service.

Anyways, we have taken all possible steps to avoid any similar hardware failure in future and even if something goes wrong with our growing server farm then we have backups to restore within a short span of time to make everything operational in short span of time.

**kanasta** · #40 (**permalink**) 24-01-2007, 01:06

RAID 1 on all Linux Servers - does that include the US so that you have which is the only server which has SSH?