Jump to content

Recommended Posts

Posted

Evening All. Just started a new position backups dont seem to be functioning correctly appearing to be down to limited storage. So im just wondering how everyone else backup these days. At my old place we used to backup to a NAS which also replicated to another NAS in different building but i dont have this option at the new site. Currently we have yosemite installed on all servers. I am running a mixture of server 2003, 2008, 2008 core and ubuntu, physical and virtual.

 

Current procedure.

Monday - Thursday - Differential of files

Friday 1,2,3,4 Full Backup of Files and system state

 

Half term - Full backup of everything.

 

Any Suggestions would be great.

Posted
Ideally locally to a nas/fileserver and then to the cloud if the internet connection was half decent.

 

Or point to point to an external site to avoid having to download critical data from the cloud. In the event of a disaster, bring your offsite target onsite and initiate recovery. Cloud is still expensive and personally not yet convinced that we should be pushing sensitive data this way (Might be just being old-fashioned with this).

 

Currently using Barracuda Networks Backup Service (point-to-point) backs up all servers without having to do major setup, emails daily reports. Backup runs incrementally to the backup server with deduplication and compression prior to replication to offsite. Expensive at £24k for 5 years but it is an insurance policy.

 

I feel that DR is not taken seriously enough in Education (imo).

Posted

I bought a ML110 G5 hp server and put 3 2tb drives into it and put them into raid0 = 6 TB of fast backup. It cost about £550 (I know it has no redundancy but it is a backup and I hated tapes)

Volume shadow copies are used day to day the 6TB is just there incase of fire. Oh and it's stored in a different building, no cloud backup but if a fire is big enough to take out the whole school then I don't think they'll mind starting from scratch.

Posted
if a fire is big enough to take out the whole school then I don't think they'll mind starting from scratch.

 

Are you sure about that?

Posted (edited)
I bought a ML110 G5 hp server and put 3 2tb drives into it and put them into raid0 = 6 TB of fast backup. It cost about £550 (I know it has no redundancy but it is a backup and I hated tapes)

 

If its RAID 0, then I'd say that no, it isn't a backup.

 

Volume shadow copies are used day to day the 6TB is just there incase of fire. Oh and it's stored in a different building, no cloud backup but if a fire is big enough to take out the whole school then I don't think they'll mind starting from scratch.

 

With all due respect, this is the craziest thing I've ever read... and I've read the source code and design spec for Vista. What about obligations for protecting exam data? Financial records? Etc? Do your SLT know and understand the implications of this "backup" process?

Edited by Roberto
Posted

Friday eve = full backup to local usb hard drive which then replicates to NAS in another building.

Mon -> = diff backup (local then replicated again).

 

As servers are virtualised, at the end of every term the VHDs are copied to an encrypted hard drive and taken off site.

Posted
I bought a ML110 G5 hp server and put 3 2tb drives into it and put them into raid0 = 6 TB of fast backup. It cost about £550 (I know it has no redundancy but it is a backup and I hated tapes)

Do you mean that this is used as a back up server or that is your only kind of DR plan?

RAID != Backup

if a fire is big enough to take out the whole school then I don't think they'll mind starting from scratch.

I would feel like its my responsibility to ensure that the network could be rebuilt as it was?

Surely its easier to get a good backup routine in place than build a new network and explain why everyone's data is gone.

  • Thanks 1
Posted
but if a fire is big enough to take out the whole school then I don't think they'll mind starting from scratch.

 

Unless there is exam material on there! I know some of the exam boards told our Exams Officer several years ago that hardware failure or loss due to fire, etc. would not be accepted as a valid excuse for not submitting anything in an electronic format that was required for examinations.

  • Thanks 1
Posted

We have:

 

Shadow copies for 'My work deleted its self' cases.

An onsite NAS unit for server failure, backup each night.

An offsite USB drive that backs up each friday and stays at home in case of fire.

SIMS and FMS data is copied to the other school I work at each night.

 

All backup data is encrypted.

Posted
if a fire is big enough to take out the whole school then I don't think they'll mind starting from scratch.

 

School fires are common enough (half a dozen or so a year, nationwide, I think I remember reading in the TES once) that there are systems in place to provide temporary replacement facilities - there are companies that specialise in rapid-deployment facilities, paid for by insurance. If your school is burnt down there's no real reason why you shouldn't be up and running again in some fashion within a week, in which case you'll need your user logins, MIS, email, etc all ready to go. Replacement servers can be got to you within 4 hours, you just need to ensure your data is ready to use on them.

  • Thanks 1
Posted (edited)

Nightly Veeam backups to disk (5x2TB in RAID5 inc hot spare) then weekly Veeam archive to LTO-5 tape... best of both worlds :)

(Using Reverse Incremental mode so the current backup job is always the main Veeam VBK file, makes it easy to archive as a point-in-time copy)

 

Ideally would be sending the Veeam backups online but ongoing cost \ hammering Internet connection for the amount of data I want uploaded doesn't make it viable at the moment.

Edited by gshaw
Posted
@dhicks - Agree with you completely there. Bizarre how some places have a less than adequate system for backup/recovery and DR. Can i ask what system you employ for this?
Posted

I've told Management how much it would cost to do this properly and they won't pay. My solution was better (just) that no solution. There are also paper copies of all the most important stuff in a third building.

 

If its RAID 0, then I'd say that no, it isn't a backup.
(yes it is, its just not redundant I used to backup to tape how is this different?)
Posted
I've told Management how much it would cost to do this properly and they won't pay. My solution was better (just) that no solution. There are also paper copies of all the most important stuff in a third building.

 

(yes it is, its just not redundant I used to backup to tape how is this different?)

 

Seriously?

You think losing a set of tapes and losing all your backups is the same thing? I have not the words, frankly.

Posted
Seriously?

You think losing a set of tapes and losing all your backups is the same thing? I have not the words, frankly.

 

It's okay though because there are "paper copies of the important stuff" - I think the Senior Leadership need to think again about how their data is managed and protected. Ditto Roberto... words fail me.

Posted

RAID0 and SATA hard drives are leaving you in a dangerous place... if one HDD packs up you lose the ability to be able to restore anyone's work from the past couple of weeks (ouch)

 

Surely there's enough £££ for 2 extra drives for RAID5 + hot spare?

Posted
Can i ask what system you employ for this?

 

We've been moving everything around over Easter (new FreeNAS-based file servers and so on), so we're still working on that - I reckon a script to snapshot our virtual machine harddrive volumes and send the diffs to a remote server should do it, along with something similar for the file servers, although we'll have to see how much data we have change every day to see if this is practical to do overnight. XenServer's "quiesced snapshot" feature looks useful for this - it would seem to let you get the virtual server into a state where you can snapshot it without having to shut it down first.

Posted

I see you're all more worried than I am:) Thanks...

 

SMT know the situation. (maybe I'll remind them) I'm only the Senior Technician here so have no real authority to demand more and can only make suggestions. The tape backup was an autoloader and spanned the backups accross tapes and they would without warning not restore some data, as a single tape had a minor error. (You could verify but the backup took over 3 days)

 

I know the raid 0 (especially on sata) isn't a sure thing but the server can't accomodate more drives, it's full. It's been running for over a year now. I need the backups so I can't reraid in 5!) Software Windows Raid at that.

Posted (edited)
Nightly Veeam backups to disk (5x2TB in RAID5 inc hot spare) then weekly Veeam archive to LTO-5 tape... best of both worlds :)

(Using Reverse Incremental mode so the current backup job is always the main Veeam VBK file, makes it easy to archive as a point-in-time copy)

 

We use Veeam here too, Mon-Thurs reverse-incrementals and Friday Full backup of our entire VM estate (about 2TB compressed, mostly thin provisioned). As for the device, it's a WD RAID 0 6TB external usb hard drive than copied to a QNAP NAS in a different building.

 

I personally think that Veeam is the best software that we have purchased. We backup at about 190MB/S and can restore entire VMs in minutes. Testing backups has never been easier!

 

Unfortunately I can't convince SMT to purchase another SAN to setup replication too!

Edited by Gardinho
Posted

Use RSYNC to a samba share on a Linux machine - it's all free.. we use it here for incremental and full backups here on a RAID 1 config.

 

Works perfect!!!! Restoration of files (if needed) is VERY easy.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...