Fun Friday: What’s the longest time you’ve seen a snapshot/checkpoint in-use for?


Userlevel 7
Badge +6

Happy Friday!

 

We know snapshots/checkpoints aren’t backups, but sometimes people don’t know this, and other times people forget to tidy up their snapshots. (Even worse when they don’t have Veeam ONE notifying them of this).

 

So the question today is, what’s the longest time you’ve seen a snapshot/checkpoint in use?

 

I’ve seen one customer environment whereby all VMs have multiple snapshots going back as far as 6 years, and counting! They’re engaged with VMware support due to the chain age and depth but they also can’t just take a backup and clear them down due to the lack of performance from the system now as a result!


22 comments

Userlevel 7
Badge +3

When working as a VMware consultant, I once spotted a virtual Exchange server with a large snapshot.

 

It was running Exchange 2013, the snapshot name was “New snap before upgrade to Exchange 2010”. Needless to say, they were lucky it was a very large datastore and no other VM's resided on it. Performance was abysmal and it took 1+ week to consolidate the snapshot according to the customer.

Userlevel 7
Badge +6

When working as a VMware consultant, I once spotted a virtual Exchange server with a large snapshot.

 

It was running Exchange 2013, the snapshot name was “New snap before upgrade to Exchange 2010”. Needless to say, they were lucky it was a very large datastore and no other VM's resided on it. Performance was abysmal and it took 1+ week to consolidate the snapshot according to the customer.

Who needs SAN upgrades when removing snapshots can give them a dramatic performance increase!

Userlevel 7
Badge +5

Wow, 6 years… unbelievable that someone would do this.

The longest time I have seen so far is something around a half year.

I have seen the warning messages in Veeam the whole time and told tthm at least twice a week to consolidate the snapshot. It took some time until they have done it…. 😎

 

Userlevel 6
Badge +1

6 years! 

Not me but I know of someone’s environment where they took snapshots each time they carried out some maintenance on their Exchange servers and never deleted them until their storage array ran out of free space. Took them about a week to consolidate whilst their Exchange servers were down. 

Userlevel 7
Badge +3

When working as a VMware consultant, I once spotted a virtual Exchange server with a large snapshot.

I had that too. A snapshot of a 2TB Exchange Server. 2-3 years old.

It wasn’t possible to consolidate, the snapshot could not be deleted. Cloning the VM was also a not successful. We had to build a new exchange server and move all data to the new server to get rid of the old one.

Userlevel 6
Badge +1

When working as a VMware consultant, I once spotted a virtual Exchange server with a large snapshot.

I had that too. A snapshot of a 2TB Exchange Server. 2-3 years old.

It wasn’t possible to consolidate, the snapshot could not be deleted. Cloning the VM was also a not successful. We had to build a new exchange server and move all data to the new server to get rid of the old one.

Ouch!

Userlevel 7
Badge +6

When working as a VMware consultant, I once spotted a virtual Exchange server with a large snapshot.

I had that too. A snapshot of a 2TB Exchange Server. 2-3 years old.

It wasn’t possible to consolidate, the snapshot could not be deleted. Cloning the VM was also a not successful. We had to build a new exchange server and move all data to the new server to get rid of the old one.

This is why snapshots are temporary 😆

Userlevel 7
Badge +6

I have seen snapshots range from weeks to over 7 years!  Yes that was fun to get rid of that one. 😂

Userlevel 7
Badge +5

I get right now a snapshot from February and it’s drives me crazy!

Veeam One help us with what at least...

 

Userlevel 7
Badge +5

Happy Friday!

 

We know snapshots/checkpoints aren’t backups, but sometimes people don’t know this, and other times people forget to tidy up their snapshots. (Even worse when they don’t have Veeam ONE notifying them of this).

 

So the question today is, what’s the longest time you’ve seen a snapshot/checkpoint in use?

 

I’ve seen one customer environment whereby all VMs have multiple snapshots going back as far as 6 years, and counting! They’re engaged with VMware support due to the chain age and depth but they also can’t just take a backup and clear them down due to the lack of performance from the system now as a result!

I just like to know how big a snapshot of 6 years can reach.

:drooling_face:

Userlevel 7
Badge +6

I get right now a snapshot from February and it’s drives me crazy!

Veeam One help us with what at least...

 

Love using Veeam ONE for this. :sunglasses:

Userlevel 7
Badge +5

Looks like Exchange server is everyone's favorite for creating snapshots...you don't know what's more fun deleting the snapshot or rolling back and destroying Exchange 🤣

 

Userlevel 7
Badge +3

I have stopped to create snapshots for exchange 4-5 years ago. A little bit of risk is needed when updating a Exchange Server. 💪😋

It‘s much more fun to do a quick rollback with veeam or to use RecoverServer Mode. 😂

.\Setup.exe /IAcceptExchangeServerLicenseTerms /Mode:RecoverServer

One time, a customer of mine has rebooted the exchange server at the same time, I was doing a scheduled maintenance (Security Update). That was a really bad idea. And he had not veeam as a backup software. RecoverServer Mode was the fastest rescue to this dilemma :)

Userlevel 7
Badge +5

I'm glad that until now I did only have to restore an Exchange for a single time. All other cases could be repaired with some effort 😅

Userlevel 7
Badge +3

I'm glad that until now I did only have to restore an Exchange for a single time. All other cases could be repaired with some effort 😅

You are lucky :)

Userlevel 7
Badge +4

I can’t remember the correct age but it was something like 1 year and also for a Exchange-server :wink: . It seems that we are not so sure about an Exchange upgrade :laughing: . As @Chris.Childerhose already mentioned, Veeam ONE is in that a live saver! Just run the active snaphots report weekly to make sure no snapshots are still active...

Userlevel 7
Badge +5

@Nico Losschaert Probably because so many things can go wrong with Exchange. And besides Internet connection, Email is one of those services where every user will start complaining when it's offline. 😐😄

By the way, for customers who don't have Veeam ONE, there's an alarm in vCenter for snapshot size; this is really useful and we often configure it.

Userlevel 3
Badge

Not about longest snapshot but fun enough to share:

One time at Veeam Support I got a problem with VBR failing to backup one VM because the VM reached the limit of snapshots. 30 snapshots in this VM. :)

Userlevel 7
Badge +6

Not about longest snapshot but fun enough to share:

One time at Veeam Support I got a problem with VBR failing to backup one VM because the VM reached the limit of snapshots. 30 snapshots in this VM. :)

As they say snapshots are not backups 😂

Userlevel 7
Badge +6

Not about longest snapshot but fun enough to share:

One time at Veeam Support I got a problem with VBR failing to backup one VM because the VM reached the limit of snapshots. 30 snapshots in this VM. :)

Would’ve loved to hear their justification for so many snapshots!

Userlevel 7
Badge +3

Wow 6 years, unbelievable, we have strict policy about snapshot now specially on production (no more than 7 days or erase). We’re more permissive on test env even due to the waiting for software integrator to have the approval from the project team.

We developped internal process to avoid snapshot, people can launch a VeeamZip or Quickbackup from enterprisemanager/Vcenter or AWS jobs.

We met 30 snapshots too due a bug between the old backup software and veeam.

Userlevel 7
Badge +7

Very bad horror stories can be read here!

I'm not sure how old the oldest snapshots I've ever seen are. But quite sure a few years old ones. To be honest most often these snapshots belongs to test VMs without any productive tasks. So most often there is no problem with this snapshots.

But I can remember some situations where I saw way too many snapshots on a single VM. Also from older vSphere versions. With vSphere 5.5 and before, ESXi created a helper-snapshot for snapshot-deletion. Then, if the deletion of the snapshots failed, it was possible that more and more snapshots remained with each deletion attempt. Very nasty!

Comment