Unsure myself on the cost impact due to not using Azure storage being an MSP we have our own.  Maybe @MicoolPaul might have some insights.

Also please don’t tag people in the community as we will see your post and try to help as best we can.  Some of those people are Veeam employees while others are just users so it is not a good idea to tag them when they may not know.

Hi @AjKumarR → Everything impacts costs in the cloud ;)

One thing I’d say is to start with a prescribed data profile stored in the cloud and build a cost model. If it meets your expectations when scaled up, that’s good. 

There are factors like number of API calls, size of data going in and recovery egress. These are all parts of the cost profile. 

Hi,

So the rule of thumb is:

Colder storage: Costs more per API call, but cheaper to store the data. Your costs will be generated based on the amount of API calls to backup your incremental data, but also consider how much it would cost if you needed to recover EVERYTHING from that backup. Then consider the likelihood of it. For example if you’ve got years of retention in archive tier, are you going to do a full restore of that data (incredibly unlikely) you’re more likely to perform file level recovery, but if it’s your primary off-site/immutable backup, then it’s far more likely you need to recover in bulk, and that recovery needs to not break the bank.

You’re going to fall foul of the 90 day minimum retention. I won’t make this exhaustive but let’s compare cool to cold (UK South LRS for reference here):

Data at rest charges: 1TB of data is £8.52pm in cool vs £3.65pm, so that’s a £4.87pm saving per TB.

Data write charges: 1TB of 1MB objects being written would be £8.71 in cool vs £14.26 in cold. So that’s a £5.55 deficit per TB written using cold.

Data read charges: 1TB of 1MB objects being read would be £0.87 in cool vs £7.92 in cold. So that’s another £7.05 deficit per TB read using cold.

Data Retrieval costs: 1TB of data retrieval is £8.92 in cool vs £24.34 in cold. So that’s another £15.42 deficit per TB read using cold.

Early deletion fees: Veeam will be keeping track of objects that are required for newer retention points so you won’t have a complete renewal of data each month for example, but you will still have blobs that are no longer required. Cool tier requires you to commit to storing your objects for 30 days, and cold tier requires you store your objects for 90 days, so anything that doesn’t require being kept for that long will incur early deletion fees. Let’s focus on a per TB figure still for simplicity, but consider your intended backup retention period. You’ve said 7 days with weekly and monthly full (Use GFS for this btw not active fulls), this means that if after 7 days you had 1TB of wasted data, you’d need to pay for it for the remaining 23 days of your minimum commitment period, so each 1TB will cost you a minimum of £8.52 in storage costs. Whilst using cold will require you keep objects for 90 days, so you’ll be committing to £10.95pm of storage costs per 1TB of data, minimum.

In summary, your backup retention likely doesn’t make sense for you to use cold.

You’ll be paying at least £10.95 per TB of data vs £8.52 per TB due to minimum retention charges

Data write charges will be more per TB

Data read charges will be more per TB

Data retrieval charges will be more per TB

I’ve focused on a 1TB baseline here so you can skew these numbers to any expected full/incremental backup sizes and get a feel for what your costs would be.

Hope it helps.