I need to upgrade an Azure Storage account from GPv1 to GPv2 because GPv1 is deprecated. Before running the migration command from the documentation, I'd like to know what preparation or best practices others recommend. The account contains a large number of blobs, and the documentation says the data and endpoints should remain unaffected. Has anyone completed this upgrade, especially at scale, and can you confirm whether the existing blobs, tables, and queues continue working normally? I'm also interested in any billing or configuration changes I should review beforehand.
4 Answers
This is generally a billing-model upgrade rather than a functional redesign. Existing resources should continue operating, but some accounts may see a noticeable price increase. Review storage, transaction, retrieval, and tier-related charges, then monitor costs after the change. The platform may also handle broad migration timelines automatically, but I wouldn’t wait without checking the current deprecation schedule and validating the account first.
The biggest thing to investigate is billing rather than data movement. GPv2 uses a different transaction pricing model and supports access tiers such as Hot, Cool, Cold, and Archive. The migration itself shouldn’t alter the blobs, but costs could change significantly for accounts with high transaction volume or a workload that isn’t suited to the default tier. Compare current usage and spend with the expected GPv2 pricing before upgrading.
I used the portal’s migration option and didn’t run into problems. Blobs, tables, and queues continued working normally, and a few spot checks before and after the change looked good. That said, results can vary by workload, so it’s worth validating your own applications and monitoring them after the upgrade.
Good to know it worked smoothly for you. I’ll make sure to test the account and application behavior rather than relying only on the migration completing successfully.
The upgrade is performed in place, so you shouldn’t need to copy the blobs into a new account. Existing data and endpoints are expected to remain available. I’d still test the process on a non-production account first, take a backup or snapshot of anything especially important, and review the account’s usage before proceeding.
That makes sense. I’ll pay particular attention to transaction counts and access patterns, since the storage capacity staying the same doesn’t necessarily mean the bill will.