How long does an EC2 vCPU quota increase usually take?
Asked by
MellowCedar42
I've reached the 8-vCPU limit for a specific AWS region and need to increase it to 16 vCPUs. I submitted a quota increase request several days ago, but I haven't received a clear response. AWS attempted to call, disconnected immediately, and then added what appeared to be an automated update to the case. Since this is only for standard m- and c-series instances, should I keep waiting on the existing request or close it and submit a new one?
Answers
Make sure you used the dedicated Service Quotas request for the specific EC2 quota, rather than opening a general support case. In many cases the quota workflow handles the increase directly, while a general ticket can take longer or be routed incorrectly. Check the request status first before submitting another one.
Reply to this answer
The instance family matters. GPU-backed families can have much stricter capacity limits and longer queues, while standard m- and c-series instances are generally easier to approve. For regular instances, this sounds more like an account or request-review delay than a capacity issue.
Reply to this answer
Account history and support status can affect quota approvals. A new individual account may receive more scrutiny even with paid support, so needing only 16 vCPUs doesn’t necessarily guarantee an immediate increase. I’d keep the current request open and add a concise update with the exact region, instance families, and business need rather than opening duplicates.
Reply to this answer
For a normal m- or c-series quota, I’d avoid closing the request unless support tells you to. Add the required capacity, region, instance types, and expected usage details to the existing request, then wait for the quota team to review it. Submitting a second request may just create duplicate work.
Reply to this answer
Add your answer
Answers are reviewed before publication.
MellowCedar42: It’s a paid-support account, but it is still relatively new and belongs to an individual. I was hoping a small increase for non-GPU machines would be routine.