Azure Cosmos DB
An Azure NoSQL database service for app development.
984 questions
Error reducing CosmosDB maximum RU usage
Michael Mehrtens (Credera)
6
Reputation points
Sometimes we need to increase our maximum RU usage in our CosmosDB instance for high load activities. When we increase our maximum RU usage, this seems to also increase the maximum RU usage that we can ever go back to.
For example, going from 15,000 to 50,000 as our maximum, we are now unable to go back down to 15,000. We get a new error.
Previously, our minimum was 10,000, but going to 100,000 raised this to 15,000, and we were unable to go back down to 10,000 again for the same reason.
Is there a way to force CosmosDB to move the maximum RU usage back to its original value?
Error updating offer for database X:
Message: {"code":"BadRequest","message":"Message: {\"Errors\":[\"The offer should have valid throughput values between 38000 and 1000000 inclusive in increments of 1000.\"]}\r\nActivityId: b3b0d951-818d-4686-9752-1dcf22d4cd48, Request URI: ..., RequestStats: \r\nRequestStartTime: 2021-10-08T13:09:44.8533054Z, RequestEndTime: 2021-10-08T13:09:44.8732958Z, Number of regions attempted:1\r\nResponseTime: 2021-10-08T13:09:44.8732958Z, StoreResult: StorePhysicalAddress: ..., LSN: 6417, GlobalCommittedLsn: 6417, PartitionKeyRangeId: , IsValid: True, StatusCode: 400, SubStatusCode: 0, RequestCharge: 0, ItemLSN: -1, SessionToken: -1#6417, UsingLocalLSN: False, TransportException: null, BELatencyMs: 13.76, ActivityId: b3b0d951-818d-4686-9752-1dcf22d4cd48, TransportRequestTimeline: [(Event: Created, StartTime: 2021-10-08T13:09:44.8533054Z, DurationMicroSecs: 0),(Event: ChannelAcquisitionStarted, StartTime: 2021-10-08T13:09:44.8533054Z, DurationMicroSecs: 0),(Event: Pipelined, StartTime: 2021-10-08T13:09:44.8533054Z, DurationMicroSecs: 0),(Event: Transit Time, StartTime: 2021-10-08T13:09:44.8533054Z, DurationMicroSecs: 19990),(Event: Received, StartTime: 2021-10-08T13:09:44.8732958Z, DurationMicroSecs: 0),(Event: Completed, StartTime: 2021-10-08T13:09:44.8732958Z, DurationMicroSecs: 0),], ResourceType: Offer, OperationType: OfferUpdateOperation\r\n, SDK: Microsoft.Azure.Documents.Common/2.14.0"}, Request URI: /offers/OT8b, RequestStats: , SDK: Microsoft.Azure.Documents.Common/2.14.0, Microsoft.Azure.Documents.Common/2.14.0, Microsoft.Azure.Documents.Common/2.14.0, Microsoft.Azure.Documents.Common/2.14.0, Microsoft.Azure.Documents.Common/2.14.0
ActivityId: 874c9eba-2a67-4563-ae0f-7b0f203fd184, Microsoft.Azure.Documents.Common/2.14.0
{count} votes
-
Anurag Sharma 17,316 Reputation points
2021-10-11T05:36:08.24+00:00 Hi @Michael Mehrtens (Credera) , welcome to Microsoft Q&A forum and apologies for the error you are facing.
Could you please try changing the throughput from Azure Portal once and let us know if you still face the same issue? Also what is the size of database/container here.
This error is coming from the Azure Portal. Size of the database
Do you mean storage size? The current size is 326 GB, but we encountered this issue when the DB was 96 GB as well. It is partitioned on ID.
Ah, is this because max RUs is set at 1000 * storage size?
https://learn.microsoft.com/en-us/azure/cosmos-db/provision-throughput-autoscale#autoscale-limits
Thanks for providing more details.
That clears the things, If we look at error message it mentions which value we need to set (starting from 37,000). And as per the official document as well, it should be close to this value based on the database size (1000 * Size). Could you please try setting it up once to 37000, it should work then? Please let us know.
Just wanted to check if you were able to try it out.
Sign in to comment