SureBackup Needs a Facelift

[bstilts]

Guys, I've been fighting SureBackup issues for weeks and the one thing I've concluded is that the development team needs to give this utility some love. I've even got the Veeam technical guys saying that this tool hasn't really changed much since version 6. The code operates more like a batch file than actual intelligent checking and processing operations. For instance, I had to up the wait times for one VM all the way to 45 minutes just to give it enough time to get through it's checking routines. When I looked at the log for one routine it ran a ping test and got back a 75% success rate. Because there happened to be one packet that got dropped it said it was going to wait an additional 2099 seconds before testing again, REALLY??? Nobody has all day to run a simple backup verification process. Please guys, spend some time on your next update going over this code and making it more efficient. Here are some possible suggestions

1.) Give the application groups the ability to skip past a failed VM. I know the original idea was to make everything in an application group required, but sometimes people want to check out some VM's in the lab environment and don't want to have to create a special job to tie it to. The application group would be great for this, but currently once one VM fails it tears down the entire job and leaves you standing there going WAIT...STOP... NOOOOOO!!!!
2.) Give the user the ability to manually select a stability checking mode. My biggest issue over the past couple weeks has been trying to get the "Stabilize by IP" checking to stop failing my SureBackup because it can't seem to get it right for one particular VM. If I could change the checking mode and just tell it to wait until the time out value expired and then move on, I could have saved myself the past couple weeks of testing and retesting and calling support and actually got some work done!
3.) Add some intelligence into the code. If it gets a 75% success rate on a ping test, then either test again immediately until you see 100% or realize that something caused the packet to drop, but that the system was still showing signs of life and move on. Why would you make the user wait an additional 2099 seconds before trying again?

Anyway, just a few ideas and I have to get back to work.

[bstilts]

Another thing that would be a nice add on would be for a simple way to select a group of VM's from a backup job and simply bypass the checking process all together. Sometimes you just want to fire up a group of VM's quickly in the lab environment and you really don't care if Veeam validates them or not, you just want the darn things to be fired up so you can check a few things off network.

[ferrus]

I have to echo some of the sentiments here.
It's a great tool and it's potential is massive, but can be slow and inflexible.

Part of the issue for us - is that we're just as interested in it's use as a sandbox environment, as it's original backup verification use.
This role is mentioned in the documentation - but not in the GUI where it's still just a 'SureBackup'. This part really hasn't been developed to its potential.

I too have waited around for 20+ minutes for a successful ping test, sitting confused while running a constant ping from my local desktop in a separate command window.
Disabling the script/heartbeat etc checks still seems to require some verification which can take a long time when someone asks for a new environment - and if there's any issues (VM tools not loading correctly on one VM) it's right back to the beginning.

An override on all of this would be great - 'just bring them online, as quickly as possible'.


After much experimentation I've found the setup that works best for us:

Configure once:
Create a linux utility VM, with Apache Guacamole installed and configured
Create a Virtual Lab with a static IP mapping to the VM

Per sandbox request:
Clone all the requested sandbox VMs - with a DC, the aforementioned linux VM and any other required infrastructure.
Backup all the cloned VMs from the previous step
Create an Application Group with the VMs from the backup, and a SureBackup with the Application Lab + VirtualLab

Now you can RDP, puTTy, or VNC to the whole environment, for as long as you like without the backups interrupting and shutting it down, and without complex network configuration each time (as long as it starts successfully).

[ferrus]

Sorry bstilts - I was still typing up my reply, before your second was posted
We agree on the 'just boot' option.

A perfect setup for me: A right-click 'On-demand Sandbox' context-menu option in the VM list, a 'Just Boot' quickly option, and a pre-installed Guacamole on the VLab VM (needing to preconfigure authentication etc would probably rule this one out!).

[DaveWatkins]

I'd also love some intelligence or settings to allow me to bring up cluster type services. At the moment anything that needs a majority quorum before becoming active always fails because the tests are done per VM as that VM comes up, so the first cluster node always reports it's failed because there aren't enough cluster nodes up. Being able to group nodes to say bring all these machines up, and once they are stable, then do the application testing for them would be awesome

[foggy]

Some of the mentioned requests have been already mentioned on the forums, but still such a detailed feedback is always valuable, so thanks everyone.