Any news on SureBackup in Veeam Backup and Replication 7.0?

[enoch]

Hi everyone,

Any news on new or better SureBackup features en Veeam Backup and Replication 7.0?

This two things i'll hope to be fixed or made possible in new version:

1) Possible to continue a SureBackup job even if one of the servers in the APP Group fails, the job should of course show as failed.
2) Possible to shutdown an Guest VM in the APP Group after verify have run on this VM*

*Some server ain't depending on each other and with 20 or more server in the SureBackup job a lot of ressources is needed on the host running these jobs.

[Gostev]

Hi Peter, it sounds like you are simply not using SureBackup Application Groups in the way they were designed to be used. Application Groups are designed to contain mandatory servers that are REQUIRED to be present and running (contrary to both of your requests) for other VMs to function normally.

Your requests mean that the specific VM you are talking about should be a linked job VM in SureBackup job, instead of being placed into the Application Group... VMs from linked jobs are processed exactly in the way you describe above (both points), even in the current version. Thanks!

[enoch]

NICE!

Learning all the time...

[enoch]

If I create an APP Group with all the mandatory servers and link an Veeam Backup job where all the mandatory servers also is included, will they then be ignored?

[enoch]

An just to ask Again, is there any news on SureBackup in 7.0 (other then SureBackup is possible on replicas)

[enoch]

If I create an APP Group with all the mandatory servers and link an Veeam Backup job where all the mandatory servers also is included, will they then be ignored?

Just tried, it looks like mandatory servers is "ignored" which is good.

I can't find it anywhere, but is it possible to get the SureBackup job to exclude some server in the linked Veeam Backup job?

An example:

In the Veeam Job: Prod Servers I have these server: DC01, DC02, WEB01, WEB02TEST, EXCHCAS01 and EXCHMB01. I create an APPGRP with DC01 and DC02. In the SureBackup job I use the APPGRP just created and link the Backup job "Prod servers". How do I get the SureBackup job to exclude on or more servers from this backup job. In my case I want to exclude WEB02TEST?

[enoch]

Just read on this and it seems standard that this is not possible. Is there any workaround for this or any changes in 7.0, it would be VERY nice to have the possibility to exclude some servers from an linked backup job

[Vitaliy S.]

I can't find it anywhere, but is it possible to get the SureBackup job to exclude some server in the linked Veeam Backup job?

That's not possible right now and I do not see this option in v7 either, but thanks for the feedback.

[enoch]

Please consider adding this minor feature ASAP

[enoch]

Hi Veeam,

Regarding the SureBackup and linked backup jobs, this is exactly what we want, but please please make the tiny fix to make it possible to exclude VM's from the linked backup job to be checked...

[enoch]

Hi Veeam,

Any news for this?

And can we soon see some new features / performance increase for SureBackup?

[Gostev]

Not until you remain the only person requesting these enhancements please understand that with hundreds of pending feature requests, it's all comes down to priorities...

By the way, there are no issues with SureBackup performance. I've seen VM boot faster off backup comparing to the same VM in production. Granted, the production VM was booting off busy older production SAN, and backup was residing on the newer SAN that was still in the test lab, so very lightly loaded. Basically, it all comes down to the speed of your backup storage - and this is something we just cannot fix, no matter how hard we try

By the way, lightning-fast SureBackup is one of many benefits that you can get by leveraging the Ultimate VM Backup Architecture!

[enoch]

I understand, by could I really be the only on who want to be able to exclude some servers from an linked backup job?. This feature should be very easy to do.

Is it just me, I havn't seen any big changed to SureBackup since version 5?

What about SureBackup performance?: Right now our physical Veeam server can read/write data with about 1000MB/sec, have 10Gbit Ethernet. SureBackup server have 1Gbit Ethernet and is on local site as the Veeam server, still it takes about 30 min. for an VM to be powered on an before checks can be run. With a lot of VM's and lot's of job this really takes very long time some times fails and other times runs without any issues. Instant recovery runs a lot faster!

[Gostev]

I understand, by could I really be the only on who want to be able to exclude some servers from an linked backup job?. This feature should be very easy to do.

You may or may not be the only one, but we have other pending features that are being requested by hundreds and thousands of users, so those features get priority. And it does not matter how simple or hard some feature is (unless it literally takes less than an hour to implement and test). Any features that require implementing new UI are not quick to implement and test by definition.

Is it just me, I havn't seen any big changed to SureBackup since version 5?

It's not just you. Why would we touch what has been polished in the course of 2 major releases, and is working perfectly, doing exactly what it says on the tin? Over the course of past 5 years, I've observed at least 3 of our major competitors deciding at some point to do a complete rewrite of what was working, and basically going out of business as the result. Never a good idea!

Instant recovery runs a lot faster!

Well, to be honest SureBackup should take exact same time, as it is based on this very Instant VM Recovery process. Basically, there is no logical explanation behind your observation, so you may want to investigate your setup with support. Perhaps, some terrible issues with the storage hosting the SureBackup VM redo logs? And using different one for Instant VM Recovery disk updates. Only that would explain the difference in performance.

[enoch]

Could be VM redo logs, but think this is on SSD disk on the VMware server used for the virtual labs. Normal speed when doing normal backup is about 5-600MB/sec to Veeam server With local storage.

I'll try to contact Veeam support monday

[mwant]

FYI I would also like to be able to exclude VM's from the linked jobs in Surebackup.

[enoch]

I really think that any news for SureBackup is dead..

No improvments the last 2-3 versions of Veeam!

[Gostev]

Peter, we prioritize all pending feature according the raw amount of requests, and until July 11st, the feature you need has only been requested once.

However, your statement about no improvements the last 2-3 versions of Veeam is incorrect! We have added a number of very major enhancements to SureBackup in the previous release, addressing all common feature requests. For example, many people wanted to populate SureBackup jobs with replicas, and we have delivered this feature. Another popular request we addressed was backup file contents verification to catch storage-level corruption as a part of SureBackup job. Finally, we have added support for Hyper-V VMs to SureBackup (which was a huge undertaking in terms of R&D effort).

These are just a few big features from the very latest release that came to my mind right away, there were more features in previous releases. Besides, there are some under-the-hood changes which simply can not be immediately seen, which yet improve the value of this feature dramatically (for example, SureBackup sessions were completely refactored in v7 based on common support issues).

In fact, knowing this from inside, I would go as far as saying, SureBackup was arguably one of the most major areas of our R&D investment in v7. Even from marketing perspective, 2 out of 7 major features that we promoted in v7 are SureBackup improvements!

[enoch]

Your right!

I still think that SureBackup was a nice feature 2 years ago, since then big improvment have lacked and a lot of small fixed and improvments haven't been seen.

This is what I think..

Still love Veeam..

[enoch]

Just this really simple thing like: "exclude VM's from the linked jobs in Surebackup"

What is the reason this havn't been fixed years ago?

[Gostev]

Only because it was requested once, while we have hundreds of pending features with dozens of requests. We must make it fair to our customer base... no matter how simple certain feature is, it will takes dev and QC time from more requested features. Does it make sense?

Posting on this forum is a great to raise the awareness of the need and start collecting requests from other users. You will hardly find any topic on this forum that has over 5 pages of requests about certain feature, where the requested feature has not been implemented

[enoch]

Hmm, not really

I fully understand that if a feature is requested a lot from many customers that this has a higher priority from Veeam.

But Features/fixes that really makes sense and takes 1 hour to get in upcoming version/patch- why not put them in?

Again i'm also pretty sure that Veeam also internally are thinking about fixes/feature that are not from the customers?

I just thin SureBackup could be SO much better if some more Development is used for this feature. SureBackup was one one the selling points to our customers 2 years ago - but no more.

[dellock6]

No more? Because all other competitors now have surebackup features? Or surereplica? I don't see them catching up on this technology, some of them just added the underlying "vPower-like" features, and not working at our same level yet...

[Gostev]

Agree with Luca. Still a selling point, because just like 2 years ago, not a single other backup vendor has anything comparable to SureBackup still...

Again i'm also pretty sure that Veeam also internally are thinking about fixes/feature that are not from the customers?

Absolutely, but they are still inspired by the customer needs which are responsible for large amount of feature requests (or complaints). In other words, customers may not be asking about a specific feature, but their request makes the need clear. To give you an analogy, after receiving many requests for a "faster horse", we are delivering a car (not exactly a faster horse, but so much better).

But Features/fixes that really makes sense and takes 1 hour to get in upcoming version/patch- why not put them in?

That is absolutely true. But unfortunately, your request does not fall into "1 hour" feature

[enoch]

I don't think we can agree about this

Again this Exlude "Bug" in linked backup job in SureBackup much be a 1 hour fix..

I just hope that SureBackup feature will be on another level soon in upcoming version/patch

[DavorS]

Well, to be honest SureBackup should take exact same time, as it is based on this very Instant VM Recovery process. Basically, there is no logical explanation behind your observation, so you may want to investigate your setup with support. Perhaps, some terrible issues with the storage hosting the SureBackup VM redo logs? And using different one for Instant VM Recovery disk updates. Only that would explain the difference in performance.

Sure backup will never be as fast as Instant VM Recovery because of built in timers (and some of them stack if server being checked has multiple roles) - the technology behind it is the same, but the logic is not. If you're doing a SureBackup for a job containing several machines that are dependant you can expect quite long times. You can ofcourse fine-tune the timers for every check but i've done it once for demo purposes and it took alot of my time so i wouldn't suggest it.

The whole idea of SureBackup is not to check your VM's fast (as it is with Instant VM Recovery), but to check it automaticly with no administrative overhead. If it means waiting a bit more than needed to make sure all services are started - so be it. I like it just the way it is now - it works like a charm. Can you optimise it even more - sure you can, but then you expose yourself to a risk of failed surebackup jobs should anything change within your infrastructure that would for example increase boot time of your machines beyond the threshold you have set up.

[enoch]

Why will SureBackup not be allmost as fast as Instant Recovery?

If I start a single server with Instant Recovery vs. starting the same server with SureBackup (not a DC) then there is a different in time.

I think speed is very important also in SureBackup. Have you tried checking 100+ VM with SureBackup daily. For us that is impossible, takes "forever"

And I also with a lot of SureBackup job see that something a SureBackup fails, if you then try to run it with Instant Recovery - no problems and sometimes you can rerun the job by SureBackup and then it is succesfull.

The Exlude "Bug" in linked backup job in SureBackup could make a different, so with don't need to boot each VM one at a time.

[DavorS]

If you start a single server - then it's the same, but you still have some "wait time" for specific checks - you can play with those timers and reduce the time needed for the job, but in my experience, it will cause you alot of administrative pain

[Vitaliy S.]

Hi Peter,

Why will SureBackup not be allmost as fast as Instant Recovery?

VM publish operation should take the same time, however keep in mind that SureBackup job also relies on VMware Tools response times. VMware Tools are required to detect Guest OS boot state and obtain config information from the VM. After that SB job proceeds with further verification tests. You can check what operation takes most of the time for the SureBackup job in its session details.

Thanks!

[bdoe]

When I setup Veeam, I kept it pretty simple. We have three hosts, each with three large file servers (several terabytes) and 8-12 other virtual machines each. I setup four jobs - one for all of vCenter except for the file servers, and one job for each file server. The full vCenter job is around 4.5TB. I like this setup since it automatically covers any new VM's I add, and I don't have to worry about jobs per host.

I've successfully created a virtual lab on one of the hosts, as well as an application group containing two domain controllers and a DHCP server. I also have a working SureBackup job, which so far is just running the application group. I'd like to expand it to everything else, but I'm not sure how to approach it. I see I can link it to jobs, but there doesn't seem to be a way to be very granular about it. Can I exclude some VM's from the job? If I go into Advanced under Link Jobs, it looks like I can set verification rules per VM, although it shows every VM, including those not part of the job I'm linking. I was thinking I could go straight to Advanced, but without linking a job I can't search for anything.

So, in short, what's the recommended approach? Do others use more jobs than I do?