What I use Veeam for

[larry]

As I rewrote my DR plans this month, I rewrote alot to use Veeam

I am using Veeam for

Quick Backup and QUICK restore. Backup Window went from 12 hours to 30 minutes. Restore went from being a project to "drag and drop". Went from having one backup a day at most to having some VM's with 4 backups everyday in a much shorter window. Restore worked every time, every test and now can be performed by a non-specialist

Veeam replication is now used for DR for a site failure. – was using our SAN to replicate the LUN to the DR SAN. This replicated a whole LUN and didn’t allow me to pick and choose which VM was the most important. Veeam use of VSS works every time on the SQL server. During a DR this required a SAN specialist, VM specialist, SQL specialist and a few hours to DR when using SAN replication. Now during a real DR it is just a right click and power on all VMs at DR site using VC. Testing now we can pick the date and time for a failover, after the test, just a quick undo fail over.

replication used for DR for a ESX server failure. – was using a powerscript to use the VC to clone but it took way longer because it didn’t use CB tracking. No report letting me know if it worked. Needed a programmer to maintain the 5 page powerscript.

replication used for DR for a SAN failure. Was using a old SAN and luck. Now I have a real plan.

Use both Backup and Replication for off site storage of backups. This alone is why I bought Veeam, our IT committee wanted to remove the risk of having our name in the paper for a lost tape in transport. Veeams backs up to disk then we write the Veeam server to tape with no load on our SAN. If the tape job fails we can now redo it during the day.

One report , one email showing 24 hours of jobs from both sites.


Thanks Veeam for a great product.

[Gostev]

Thanks Larry for sharing your experiences, I think it will be good idea to stick this thread for a while so that other people can pick up some ideas from this, as well as share more use cases.

I really want to mention this other use case you posted too - I find it very, very creative!
Really good use of otherwise useless local ESX storage: Using Veeam for DR of the SAN

I know you have this bullet in the list above, but from there it is not clear/obvious how exactly you do this

[Exrace]

This is perfect.
I hope you don't mind but I am going to use much of this in a doc I am writing up for a proposal to use Veeam at the hospital I work at.

[larry]

It is ok becuase I cut it from the docs I had to write.

Under Veeam replication is now used for DR for a site failure
I should of added.

I have two main sites, each are the DR of the other site. Both sites have production every day. My new DR plan I want to be able to mount the replicated DR VMs at will with little work. I am placing extra nics in all my production ESX servers at both sites. All ESX server will have their NIC setup the same with the same name. This way I can restore a VM in the DR site and the VM’s NICs will map to their normal NIC which is now on a DR network. The DR network would have no interaction with my production network. This will save me days of work to do a full test and I will be able to perform it during the day.

this has now been tested and worked as planned.

I bought Veeam for backup but it is the best DR tool I have ever used.

[jgremillion]

We do the same thing here. It handles about 10 TB a night pretty good while staying in our 12 hr window. I also use Veeam Monitor a lot to keep track of performance during the day and review performance of VMs during our backup window. No problem at all with Veeam Backup (except we keep running out of storage space as I keep adding VMs ).

Good stuff this Veeam is!

[Exrace]

The hard part is getting others on the team to see this as a better method then old style backups and "pray it works" restores.
Agent restores using a tool like Commvault has its pluses when restoring files on a users day to day file shares when they delete files by mistake etc but I have yet see it restore a complete server system without major headaches and elapsed time.

When 5.0 comes out it will make DR even easier.
Thanks for posting your details.

[applc]

Tried posting earlier and failed, so I hope this isn't duplicated...

For Site DR, are you using strictly Veeam Replication to do this? Or are you using it in combination with VMware Site Recovery Manager, or snapshot/replication tools that came with your hardware SAN? If Veeam does it all, then I know what product I'm buying next!

Thanks for any insight you can provide.

[larry]

I use Veeam to do it all now. In each DR site I have all the same Nic names on all the ESX servers as the production ESX servers. This way when Veeam replicates, I can just turn on a AD server and what ever other servers I want to test. I plug PCs into the DR network, all the same IPs as the production network and test. If this wasa real DR I would just jumper between the DR switch and the production network at the DR site, change routes on the WAN to route old production IP addreess to DR site, done. ( most likley would reassign a few nics in VC to the DR lan ) I tested this a few times this month, all worked as planned.

[icebun]

Hey Larry,

How how you configured your Veeam server?

Is it Physical or a VM and are using the SAN attached mode?

[larry]

Hardware ( Both boxes are the same )
Systemax Intel Xeon Dual core E5404P 2.00
4 SATA Seagate ES 1TB drives on internal intel raid card.
4 gb ram

[TrevorBell]

Real life test:

call at 13:10 yesterday to attend the boardroom ( high level meeting ) director " this excel spreadsheet is curropt i need one from our backup "
im thinking Hmm oki how long should i quote as recovery to get it back onto his laptop ??? i said " yesterdays backup will take around 7 minutes to restore this file" All other boardmembers looked like i was talkign crap!
Went back to IT Room mounted last nights .vbk around 2.04TB and copied file directly to his H: drive shared. Walked back to boardroom and said your file has been replaced please try it.
Director was amazed at the speed!

Total time to restore start to finish 3 minutes!

[Gostev]

Nice

[TonioRoffo]

Don't get me wrong - I love veeam backup & replication and use it for every virtual environment I have a say in!

However... You can beat that 7 mins sir! I'm assuming you are running windows? In this case, the best restore is... no restore at all. For file servers I always configure windows shadow copies on the volume containing the share, making at least 2 copies a day. Then finetune the used space so you have a few days worth of copies. (usually, before the start of the work day, and around noon). You could have right clicked the excel file on your server and restored an older version or... the director could have right clicked his file and check for "previous versions" himself, if he has the right OS...

This added level of "backup", especially against user errors, is so easy to setup, and easy to use that I consider it mandatory for windows servers.

[Gostev]

I am afraid shadow copies won't help you at all if that Excel file got corrupted due to underlying storage corruption. Shadow copies are only good to revert unintensional changes of documents (by users, viruses, etc.), think of them as source control/history system for documents and files.

[TonioRoffo]

That of course, is true! I was assuming the corruption of the file was user error.

[Bunce]

User error is far more common than storage corruption. Shadow copies are a must - and so easy to setup.

[mellerbeck]

Veeam replication is now used for DR for a site failure. – was using our SAN to replicate the LUN to the DR SAN. This replicated a whole LUN and didn’t allow me to pick and choose which VM was the most important. Veeam use of VSS works every time on the SQL server. During a DR this required a SAN specialist, VM specialist, SQL specialist and a few hours to DR when using SAN replication. Now during a real DR it is just a right click and power on all VMs at DR site using VC. Testing now we can pick the date and time for a failover, after the test, just a quick undo fail over.

Awesome write up! That is where I am trying to get to! I have lots of questions of course! First, what is the speed and distance between sites. Do you use WAN acceleration. Do you 'pull' VM's to your DR site? Do you use some type of log shipping to sync up your veeam DB?

How do you handle sql log files with veeam. Do you write them to the VM and then snapshot the whole thing for a backup?

Thanks for any thoughts!

[larry]

This post has my WAN setup.
This post has a zip file which has a word doc showing my DR setup.

I pull all backups and replications. This way the remote site always has all the information. My WAN is 100 meg ( 90 meg used by Veeam at night ) with 0 < 5 ms ping. ( 20 miles )

We DR one SQL by replicating at night but sending the logs over the WAN real time. This way we can rebuild quickly without effecting the database with snapshots during peek time. Logs are on their own disk.

Other SQL servers are just replicated using vss normally.

[Gostev]

We have put the information from this topic into the following document
http://www.veeam.com/go/7-reasons-to-ch ... are-backup

Thanks again Larry for sharing this with the community!

I am going to unsticky the thread since this made it to www.veeam.com front page anyway