search

MicrosoftDocs / azure-docs

Open source documentation of Microsoft Azure
https://docs.microsoft.com/azure
Creative Commons Attribution 4.0 International
10.27k stars 21.45k forks source link

vCore change compute resources #24881

Closed HeliosLHD closed 5 years ago

HeliosLHD commented 5 years ago

The description here doesn't seem to be what I've seen in practice, a Gen4 1vCore 850GB DB can scale to 8 vCore in about 3 minutes, not 3 hours. Isn't the vCore model using a different method to scale compute resources versus a DTU (which I can also confirm takes hours).

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

Mike-Ubezzi-MSFT commented 5 years ago

@HeliosLHD Thank you for detailing this and bringing it to our attention. I realize this can be frustrating, and although there is no immediate documentation needing to be addressed, I would really like to get you to Azure Support to investigate this, as a means of investigating and identifying an underlying issue that is preventing this functionality from meeting SLA. If you have a support plan, please open a support ticket from the Azure Portal. If you don't have an Azure Support Plan, please send us your Subscription ID to AzCommunity at microsoft.com and I will return instructions to create a support request and have this specific issue investigated. Thank you!

HeliosLHD commented 5 years ago

@Mike-Ubezzi-MSFT

I don't think this an SLA issue, I'm just curious if the description of scaling process for vCore compute resources is incorrect. I was reviewing available options for us recently and testing the new vCore model on some test DBs, and noted scaling is happening much faster than this description, which is great!!! I just want to make sure I understand all the differences or see if I'm missing something.

I can create a support ticket under or Azure Support Plan, but it's more of a question with the documentation and architecture than an issue to be resolved.

Thanks, Roger

Mike-Ubezzi-MSFT commented 5 years ago

@HeliosLHD Cool! I was anticipating the worse but, it appears you are providing some good feedback. A support ticket is not necessary here but, if there are specific sections of the document that you would like clarification upon, this channel is the appropriate means to have doc related issues addressed. If you detail your questions more explicitly, we can investigate and provide clarification.

HeliosLHD commented 5 years ago

@Mike-Ubezzi-MSFT

The section of this on vCore looks like it was copied from DTU, but Azure does not behave this way in practice with SQL vCore scaling, it completes in a few minutes and not 3 hours, so I expect the description of the process for vCore scaling is incorrect.

"vCore-based purchasing model: Change compute resources After initially picking the number of vCores, you can scale a single database up or down dynamically based on actual experience using the Azure portal, Transact-SQL, PowerShell, the Azure CLI, or the REST API.

Changing the service tier and/or compute size of a database creates a replica of the original database at the new compute size, and then switches connections over to the replica. No data is lost during this process but during the brief moment when we switch over to the replica, connections to the database are disabled, so some transactions in flight may be rolled back. The length of time for the switch-over varies, but is generally less than 30 seconds 99% of the time. If there are large numbers of transactions in flight at the moment connections are disabled, the length of time for the switch-over may be longer.

The duration of the entire scale-up process depends on both the size and service tier of the database before and after the change. For example, a 250-GB database that is changing to, from, or within a General Purpose service tier, should complete within six hours. For a database the same size that is changing compute sizes within the Business Critical service tier, the scale-up should complete within three hours."

Mike-Ubezzi-MSFT commented 5 years ago

@HeliosLHD Thank you for detailing the specifics as it helps to bring accuracy to the documentation. I have assigned the issue to the content author to evaluate and update as appropriate.

Mike-Ubezzi-MSFT commented 5 years ago

@julieMSFT Can you review this and evaluate the document for possible improvement.

julieMSFT commented 5 years ago

@HeliosLHD Hello, Thank you for your input. I will be updating the content to say the following:

"The duration of the entire scale-up process generally depends on both the size and service tier of the database before and after the change. For example, any size database that is changing the compute size within the General Purpose service tier should complete within several minutes On the other hand, the latency to change the compute size within the Business Critical tier is generally 90 minutes or less per 100 GB."

julieMSFT commented 5 years ago

@HeliosLHD I hope this addressed your issue. I will proceed with closing this issue. Please feel free to contact us if we can be of further assistance.

julieMSFT commented 5 years ago

PR #67120 - https://github.com/MicrosoftDocs/azure-docs-pr/pull/67120

please-close