Cosmos DB data replication with a second region stuck at "Updating" now for over 3 days
We have four subscriptions on a single Azure tenant, all have separate cosmos accounts and are in the UK South region.
We wanted to simply try out global data replication using a second ("UK West") read region on our sandbox subscription before turning that on on our production subscription, if there were no issues. However at the top of the page it says "Updating" for the last four days. It also says "Deleting" next to the "UK West" read region. There is nothing we can do as everything non this page seems disabled now. We have also found we are having problems experimenting with other Cosmos related features.   
Can you please investigate this asap.  
What we are finding most surprising is that we have had to purchase a Developer support plan just to make contact with you so we can notify you of your own issue. We would have thought that any cloud hosting platform would have automatically resolved such a crash sooner than this.
Azure Cosmos DB
- 
							
					
	
						Amira Bedhiafi • 39,646 Reputation points • Volunteer Moderator2025-10-28T18:36:44.8833333+00:00 Hello Marty !
Thank you for posting on Microsoft Learn Q&A.
I think if you have any other account change that overlaps, the region update can get stuck in updating or deleting. Cosmos explicitly doesn't allow changing regions while you change other properties and the portal will keep spinning even though nothing is progressing.
You have to push a region only update or revert to single region to clear the stuck state.
Do also rule out a coincident regional incident before you start.)
https://docs.azure.cn/en-us/cosmos-db/how-to-manage-database-account
 - 
							
					Swapnesh Panchal • 750 Reputation points • Microsoft External Staff • Moderator
2025-10-28T18:46:36.2366667+00:00 Hi Marty Barr,
Welcome to the Microsoft Q&A and thank you for posting your questions here.
You were adding UK West as a read region. Cosmos DB serializes region changes; if another account property was changed around the same time, the region operation can get stuck showing “Updating/Deleting.”-  Push a region-only update (no other changes) or revert to single region via PowerShell/CLI:
-  PS: 
Update-AzCosmosDBAccountRegion -OperationType Add/Remove -Location "UK West" -  CLI: 
az cosmosdb update --locations ... 
 -  PS: 
 - Make sure there are no Resource Locks / Azure Policies blocking updates.
 - Check Service Health for UK South/UK West.
 
If it’s still stuck after a single retry (or >2 hours), please share required details in private message.
 -  Push a region-only update (no other changes) or revert to single region via PowerShell/CLI:
 - 
							
					Marty Barr • 0 Reputation points
2025-10-28T20:01:34.1866667+00:00 if another account property was changed around the same time, the region operation can get stuck showing “Updating/Deleting.”
Nothing else was changed during the same time so I'm not sure why I need to start invoking PS/CLI commands. The portal seems very buggy. I simple added the UK West region and clicked save. Please investigate logs on your side and confirm what other changes were made on my account at that time. I've submitted my subscription and resource information to you privately FYI.
 - 
							
					VRISHABHANATH PATIL • 1,380 Reputation points • Microsoft External Staff • Moderator
2025-10-29T07:21:44.1466667+00:00 Hi Marty Barr
Could you please check the solution provided by Amira Bedhiafi in answer section and share your feedback, so we can troubleshoot further.
Thanks,
Vrishabh - 
							
					Marty Barr • 0 Reputation points
2025-10-29T08:13:20.05+00:00 This is not relevant as I believe I did not make any other changes to my account when I simply requested a second cosmos read region to be added. I don't think also I should be investigating an obvious bug with Cosmos data replication by running CLI or PS comamnds.
 - 
							
					Marty Barr • 0 Reputation points
2025-10-29T08:13:59.7+00:00 That solution is irrelevant. I've replied to it.
 - 
							
					Swapnesh Panchal • 750 Reputation points • Microsoft External Staff • Moderator
2025-10-30T01:24:57.75+00:00 Hi Marty Barr
Thank you for your feedback. Our team is currently working on this and will get back to you shortly. - 
							
					Marty Barr • 0 Reputation points
2025-10-30T15:21:49.0033333+00:00 It's been now 5 days with Cosmos global data replication stuck at "Updating" on my Cosmos account. Is there any way this could be fixed so we could use data replication or should we look into moving hosting providers?
 - 
							
					Marty Barr • 0 Reputation points
2025-10-31T09:21:35.8833333+00:00 This is what see at the Settings -> Replicate data globally on our Cosmos service. So it is clearly broken and unusable right now. We can't remove "UK West" read region or add a different one. Can you please fix this soon?
 - 
							
					Kalyani Kondavaradala • 3,310 Reputation points • Microsoft External Staff • Moderator
2025-10-31T13:13:00.0133333+00:00 Hi Marty Barr,
Thanks for your patience! I verified the backend logs and could see you were trying to add the region on 2025-10-24T20:33:33.9730316Z and the request failed with below error
Message: {"Errors":["Your account is currently configured with a total throughput limit of 2800 RU/s. This operation failed because it would have increased the total throughput to 5600 RU/s. See https://aka.ms/cosmos-tp-limit for more information."]}
please increase RU's to overcome the above error.
and regarding the UI issue "portal stuck on updating" i am checking internally with product group team will get back to you as soon as i hear back from them.
I appreciate your patience on this topic, do let me know incase of further queries, I would be happy to assist you.
 - 
							
					Kalyani Kondavaradala • 3,310 Reputation points • Microsoft External Staff • Moderator
2025-11-03T08:21:55.4533333+00:00 Hi Marty Barr,
Have you attempted to increase the Request Units (RUs)? Based on the logs, it appears that the server is online.
Is the portal still in the updating state? If so, this issue might be related to the Azure portal itself. To investigate further, could you please provide the following details?
- Portal URL
 - Portal session ID (can be captured using Ctrl + Alt + A diagnostics output or Ctrl + Alt + D diagnostics popup)
 - Network trace/HAR file
 
 - 
							
					Marty Barr • 0 Reputation points
2025-11-03T09:23:40.4233333+00:00 We haven't yet attempted the RU increase. I can do that today though. We just wanted to experiment with global data replication on Sandbox for a few days initially before switching that off there and switching it on in our production environment. However based on the length of time it has taken to resolve this issue (over a week now with still no fix in sight) we don't feel like we want to waste any money given that the increase in RU will start incurring a charge on our Sandbox environment right away.
Yes it is still in an "Updating" State as you can see form the screen shot.
I will send you those details via private message.
 - 
							
					Kalyani Kondavaradala • 3,310 Reputation points • Microsoft External Staff • Moderator
2025-11-03T10:58:37.4466667+00:00 Hi Marty Barr,
Could you please check your private messages and follow the HAR file download process shared there? We need these details to investigate further.
From our backend, the server appears to be online, which suggests the issue might be related to the portal UI. Providing the HAR file will help us analyze the behavior and assist you further.
 - 
							
					Marty Barr • 0 Reputation points
2025-11-03T12:44:00.99+00:00 There is nothing in my private messages on this thread where you have described how to follow the HAR file download process. In addition, the private messages seem broken. It is showing the author of messages that I posted as authored from MS engineers and vice versa it's showing messages from you as having been authored by me :-). Place describe the HAR file download process here or by private message and I will be happy to assist you fix your own bugs.
 - 
							
					Kalyani Kondavaradala • 3,310 Reputation points • Microsoft External Staff • Moderator
2025-11-03T13:17:33.82+00:00 Hi Marty Barr,
Can you please go through the below process and provide us the HAR file.
- Open Developer Tools: Press F12 (or Ctrl + Shift + I on Windows/Linux) in your browser.
 - Navigate to the Network tab.
 - Ensure Preserve log is checked (this prevents the log from clearing on page navigation).
 - Reproduce the Issue: Clear any existing logs in the Network tab, then reproduce the specific page load or action where the problem occurs.
 -  Find the ID:
-  Look for a major request to a Microsoft domain (e.g., 
portal.azure.com). - Select the request, go to the Headers tab, and look for response headers like:
 -  
x-ms-client-request-id -  
x-correlation-id -  
x-ms-session-id -  
activityid 
 -  Look for a major request to a Microsoft domain (e.g., 
 - Export the HAR: Right-click anywhere in the list of requests and select Save all as HAR (or Export HAR)
 
I have provided these steps over Private Message, there might be glitch that's the reason its coming vice versa.
 - 
							
					
	
						Carlos Jesús Rojas Gutiérrez • 425 Reputation points2025-11-04T03:18:58.5033333+00:00 Este problema con Azure Cosmos DB donde la replicación de una segunda región queda bloqueada en estado “Actualizando” o “Eliminando” por varios días es un comportamiento conocido que puede deberse a una operación atascada o conflicto interno en la cuenta.
Causas comunes del bloqueo en replicación de regiones
-  Cambios simultáneos en la cuenta
- Cosmos DB serializa las operaciones de configuración, por lo que si se intentó modificar otra propiedad (como throughput, claves, etc.) al mismo tiempo que se agregaba la región, la operación puede quedar atascada.
 
Políticas de acceso, bloqueos de recursos o errores de replicación regional pueden impedir la finalización. 
Soluciones recomendadas por Microsoft
1. Forzar una actualización de región vía PowerShell o CLI
Ejecuta una operación de actualización que solo modifique las regiones, sin tocar otras propiedades:
PowerShell:
powershell
Update-AzCosmosDBAccountRegion -ResourceGroupName "<grupo>" ` -Name "<nombreCuenta>" ` -Location "UK West" ` -OperationType RemoveCLI:
bash
az cosmosdb update --name <nombreCuenta> --resource-group <grupo> --locations regionName="UK South" failoverPriority=0 isZoneRedundant=FalseEsto puede desbloquear el estado atascado.
2. Revertir a una sola región
Si no necesitas la segunda región de inmediato, puedes revertir temporalmente a una sola región para limpiar el estado.
3. Verificar incidentes regionales
Asegúrate de que no haya incidentes activos en UK South o UK West en Azure Status.
4. Contactar soporte con detalles técnicos
Si ya tienes un plan de soporte, proporciona:
Nombre de la cuenta
Región afectada Hora aproximada del cambio ID de suscripción Capturas de pantalla del portalEsto acelera la investigación del backend.Este problema con Azure Cosmos DB donde la replicación de una segunda región queda bloqueada en estado “Actualizando” o “Eliminando” por varios días es un comportamiento conocido que puede deberse a una operación atascada o conflicto interno en la cuenta.
Causas comunes del bloqueo en replicación de regiones
Cambios simultáneos en la cuenta
Cosmos DB serializa las operaciones de configuración, por lo que si se intentó modificar otra propiedad (como throughput, claves, etc.) al mismo tiempo que se agregaba la región, la operación puede quedar atascada.
Estado inconsistente en el portal
El portal puede mostrar “Actualizando” o “Eliminando” indefinidamente aunque no haya progreso real. **Restricciones de red o políticas** Políticas de acceso, bloqueos de recursos o errores de replicación regional pueden impedir la finalización.Soluciones recomendadas por Microsoft
1. Forzar una actualización de región vía PowerShell o CLI
Ejecuta una operación de actualización que solo modifique las regiones, sin tocar otras propiedades:
PowerShell:
powershell
Update-AzCosmosDBAccountRegion -ResourceGroupName "<grupo>" ` -Name "<nombreCuenta>" ` -Location "UK West" ` -OperationType RemoveCLI:
bash
az cosmosdb update --name <nombreCuenta> --resource-group <grupo> --locations regionName="UK South" failoverPriority=0 isZoneRedundant=FalseEsto puede desbloquear el estado atascado.
2. Revertir a una sola región
Si no necesitas la segunda región de inmediato, puedes revertir temporalmente a una sola región para limpiar el estado.
3. Verificar incidentes regionales
Asegúrate de que no haya incidentes activos en UK South o UK West en Azure Status.
4. Contactar soporte con detalles técnicos
Si ya tienes un plan de soporte, proporciona:
Nombre de la cuenta
Región afectada Hora aproximada del cambio ID de suscripción Capturas de pantalla del portalEsto acelera la investigación del backend.
 -  Cambios simultáneos en la cuenta
 
Sign in to comment