-
- Expert
- Posts: 140
- Liked: 18 times
- Joined: Feb 08, 2018 3:47 am
- Full Name: Kazz Beck
- Contact:
Capacity Tier clean up script
Hello,
I have a backup job with one year retention policy, the job runs daily with a full backup once a week.
SOBR is the repository for this job with 7 days of restores kept on the performance tier and 365 restores kept on the capacity tier.
I am offloading as soon as the backup job finishes.
Looking at my performance tier, I am seeing restore files older than 90 days, I am also seeing files older than 7 days are reduced in size (deflated) as they have been offloaded.
It's my understanding that Veeam will not physically delete offloaded files (older than 7 days) from the performance tier per job retention being set to 365 days.
If I were to restore anything older than 7 days, I have to rely on the capacity tier as what I have on the performance tier could not be used.
Wondering if anyone had to deal with a clean up of the performance tier and already wrote a PowerShell script that takes care of removing orphaned files from the performance tier?
Thank you
I have a backup job with one year retention policy, the job runs daily with a full backup once a week.
SOBR is the repository for this job with 7 days of restores kept on the performance tier and 365 restores kept on the capacity tier.
I am offloading as soon as the backup job finishes.
Looking at my performance tier, I am seeing restore files older than 90 days, I am also seeing files older than 7 days are reduced in size (deflated) as they have been offloaded.
It's my understanding that Veeam will not physically delete offloaded files (older than 7 days) from the performance tier per job retention being set to 365 days.
If I were to restore anything older than 7 days, I have to rely on the capacity tier as what I have on the performance tier could not be used.
Wondering if anyone had to deal with a clean up of the performance tier and already wrote a PowerShell script that takes care of removing orphaned files from the performance tier?
Thank you
-
- Chief Product Officer
- Posts: 31419
- Liked: 6992 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland
- Contact:
Re: Capacity Tier clean up script
Hello, you should not be cleaning up anything manually from the performance tier manually or with scripts. Those stub files are not "orphaned", they will be deleted by the retention policy after 365 days. Thanks!
-
- Expert
- Posts: 140
- Liked: 18 times
- Joined: Feb 08, 2018 3:47 am
- Full Name: Kazz Beck
- Contact:
Re: Capacity Tier clean up script
Anton, thanks for a quick response.
Is there a setting on a job or repository I am missing that would remove stub files older than 7 days from the performance tier?
If I lose my performance tier due to a hardware failure, I would still be able to restore using the capacity tier.
I guess I would like to understand the reasoning for leaving the stub files alone, they seem to add up in size over time especially if I am backing up over 100 VMs
Is there a setting on a job or repository I am missing that would remove stub files older than 7 days from the performance tier?
If I lose my performance tier due to a hardware failure, I would still be able to restore using the capacity tier.
I guess I would like to understand the reasoning for leaving the stub files alone, they seem to add up in size over time especially if I am backing up over 100 VMs
-
- Chief Product Officer
- Posts: 31419
- Liked: 6992 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland
- Contact:
Re: Capacity Tier clean up script
They hold a local copy of the metadata for each offloaded restore point, so we don't have to query this metadata directly from object storage each time we need it. This helps both performance and costs (as most cloud object storage providers will charge you for reading your data back).
-
- Expert
- Posts: 140
- Liked: 18 times
- Joined: Feb 08, 2018 3:47 am
- Full Name: Kazz Beck
- Contact:
Re: Capacity Tier clean up script
Got it. So if I am more concerned about free space on my performance tier than queries to object storage and performance loss, I can delete the stubs, but it sounds like I am going against official Veeam recommendation.
-
- Chief Product Officer
- Posts: 31419
- Liked: 6992 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland
- Contact:
Re: Capacity Tier clean up script
If you delete stubs, some operations will start failing until you execute a SOBR rescan, which will recreate missing stub by downloading data from object storage.
-
- Expert
- Posts: 140
- Liked: 18 times
- Joined: Feb 08, 2018 3:47 am
- Full Name: Kazz Beck
- Contact:
Re: Capacity Tier clean up script
Sounds like stubs need to stay.
While we are on the same subject, I have another job with a 30 day retention in capacity and 3 days retention in performance. Full is taken 3 times a week and offload is happening as soon as backup is finished. While looking at the performance tier I am seeing files way older than 60 days for many VMs, but there is only 4 instances of restore points for each vm that are older than 30. I suspect there was a failed back up job at the time and files are truly orphaned as they are outside of the retention window.
Here is an example
VM1-12-15-2022.VBK
VM1-12-14-2022.VIB
VM1-12-13-2022.VIB
VM1-12-12-2022.VBK
VM1-12-11-2022.VIB
VM1-12-10-2022.VIB
........ (to keep this post shorter I am gonna skip listing all files, but they do exist)
VM1-11-14-2022.VIB
VM1-12-13-2022.VIB
VM1-11-12-2022.VBK
VM1-11-11-2022.VIB
I do understand the possibility of getting 34 restore points since I am not taking full backups daily, but seeing restore points older than 35 days is something that should not happen.
I have 4 restore points for VM1 older than 34 days. They are not listed in the DB for the job.
VM1-09-05-2022.VBK
VM1-09-04-2022.VIB
VM1-09-03-2022.VIB
VM1-09-02-2022.VIB
Maybe what I should be asking for is a script that looks in Veeam DB for a specific job, gets a list of all restore point names then compares names stored in the DB to the list of the actual files on the performance tier and provides a report or extra or missing files for each VM ?
Thanks
While we are on the same subject, I have another job with a 30 day retention in capacity and 3 days retention in performance. Full is taken 3 times a week and offload is happening as soon as backup is finished. While looking at the performance tier I am seeing files way older than 60 days for many VMs, but there is only 4 instances of restore points for each vm that are older than 30. I suspect there was a failed back up job at the time and files are truly orphaned as they are outside of the retention window.
Here is an example
VM1-12-15-2022.VBK
VM1-12-14-2022.VIB
VM1-12-13-2022.VIB
VM1-12-12-2022.VBK
VM1-12-11-2022.VIB
VM1-12-10-2022.VIB
........ (to keep this post shorter I am gonna skip listing all files, but they do exist)
VM1-11-14-2022.VIB
VM1-12-13-2022.VIB
VM1-11-12-2022.VBK
VM1-11-11-2022.VIB
I do understand the possibility of getting 34 restore points since I am not taking full backups daily, but seeing restore points older than 35 days is something that should not happen.
I have 4 restore points for VM1 older than 34 days. They are not listed in the DB for the job.
VM1-09-05-2022.VBK
VM1-09-04-2022.VIB
VM1-09-03-2022.VIB
VM1-09-02-2022.VIB
Maybe what I should be asking for is a script that looks in Veeam DB for a specific job, gets a list of all restore point names then compares names stored in the DB to the list of the actual files on the performance tier and provides a report or extra or missing files for each VM ?
Thanks
-
- Expert
- Posts: 140
- Liked: 18 times
- Joined: Feb 08, 2018 3:47 am
- Full Name: Kazz Beck
- Contact:
Re: Capacity Tier clean up script
HannesK mentioned capacity tier with immutability and REFS on the performance tier in an old post.
powershell-f26/cleanup-orphaned-vib-files-t70775.html
That is exactly how we are set up. Can someone elaborate on the issue with more details?
powershell-f26/cleanup-orphaned-vib-files-t70775.html
That is exactly how we are set up. Can someone elaborate on the issue with more details?
-
- Chief Product Officer
- Posts: 31419
- Liked: 6992 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland
- Contact:
Re: Capacity Tier clean up script
Kazz, kindly don't mix all sorts of different questions into the single topic, as this is unmanageable.
If you need a different script now, please create a new topic and hopefully some other community member who is able to help with the particular query will pick it up. For example, I have engaged in this topic despite my little PowerShell knowledge because I knew I would only need to explain why you do NOT need the requested script
And if you want to learn more about that issue, then please ask in the actual topic where the issue is being discussed.
If you need a different script now, please create a new topic and hopefully some other community member who is able to help with the particular query will pick it up. For example, I have engaged in this topic despite my little PowerShell knowledge because I knew I would only need to explain why you do NOT need the requested script
![Very Happy :D](https://forums.veeam.com/images/smilies/icon_biggrin.gif)
And if you want to learn more about that issue, then please ask in the actual topic where the issue is being discussed.
-
- Expert
- Posts: 140
- Liked: 18 times
- Joined: Feb 08, 2018 3:47 am
- Full Name: Kazz Beck
- Contact:
Who is online
Users browsing this forum: No registered users and 2 guests