Scaling up Synapse Dedicated Pools taking a long time

Question

Scaling up Synapse Dedicated Pools taking a long time

Anthony Perkins 0

Scaling up Synapse Dedicated Pools is taking more than an hour

Swapnesh Panchal 5 Reputation points Microsoft External Staff Moderator

2025-08-26T18:14:29.7033333+00:00
Hi Anthony Perkins,

Welcome to the Microsoft Q&A Platform

It sounds like you're facing delays when scaling up your Synapse Dedicated SQL Pools, with operations taking more than expected.
Here are a few quick checks:

Session Cleanup: The system must terminate active sessions and roll back any running transactions before scaling begins.

Compute Reconfiguration: It detaches current compute nodes, provisions new ones, and reattaches storage.

Background Load: If the system is under pressure (e.g., peak hours or heavy ETL), scaling can be delayed.

Current Workload: Ensure no heavy ETL jobs, BI tools, or long-running queries are active during scale.

Existing Transactions: Use DMVs to identify sessions or requests that are still running and may be delaying rollback.

Pool Status: Use Azure CLI to check if the pool is stuck in Scaling or transitioning to Online.
az synapse sql pool show \
--name  \
--workspace-name  \
--resource-group  \
--query "{name:name, status:status, sku:sku.name}"

Operation Progress: Run this SQL query to track control plane operations.
select * from sys.dm_operation_status where resource_type_desc = 'Database' and major_resource_id = 'YourDatabaseName';

Suggestions to Improve Scaling:

Drain Workload Before Scaling: Stop pipelines and apps to prevent rollback delays.

Scale During Off-Peak Hours: Avoid batch windows or peak business times.

Tune Resource Classes: Avoid using larger for routine jobs to reduce concurrency pressure.

Avoid Overlapping Operations: Don’t trigger multiple scale/pause/resume actions at once.

If you're still facing issues, please help us troubleshoot by sharing the following details:

When did the scale attempt start? (Include date, time, and time zone)

What DWU level are you scaling from and to?

Which method was used to initiate the scale? (e.g. Portal, CLI)

Were any workloads active at the time? (ETL jobs, BI tools, pipelines, etc.)

If available, please also share any relevant logs or screenshots to help us investigate further.

If you have any further questions, feel free to respond in this same thread.
Swapnesh Panchal 5 Reputation points Microsoft External Staff Moderator

2025-08-28T11:42:15.5666667+00:00
Hi Anthony Perkins,

Thanks for your question! Scaling delays in Synapse Dedicated SQL Pools can result from session cleanup, compute reconfiguration, or active workloads. To improve scaling speed:

Drain workloads before scaling.

Scale during off-peak hours.

Use Azure CLI or PowerShell to monitor status.

Avoid overlapping operations.

Also, consider reviewing your distribution strategy and resource class usage. If you're scaling from DW300 to DW1000 and have active jobs (e.g., on-prem data copies), that could explain the delay.

If you have any further questions, feel free to respond in this same thread.

1 answer

Your answer

Swapnesh Panchal 5 Reputation points Microsoft External Staff Moderator

2025-08-28T11:42:15.5666667+00:00

Hi Anthony Perkins,

Thanks for your question! Scaling delays in Synapse Dedicated SQL Pools can result from session cleanup, compute reconfiguration, or active workloads. To improve scaling speed:

Drain workloads before scaling.

Scale during off-peak hours.

Use Azure CLI or PowerShell to monitor status.

Avoid overlapping operations.

Also, consider reviewing your distribution strategy and resource class usage. If you're scaling from DW300 to DW1000 and have active jobs (e.g., on-prem data copies), that could explain the delay.

If you have any further questions, feel free to respond in this same thread.

Answer 1

Anthony Perkins 0

Started today at 8:34 and finally completed at 9:26 GMT-0500 (Central Daylight Time)
From 300DWU to 1000DWU
Portal
There were a couple of jobs copying data from on-prem
See attached from the Activity Log

Swapnesh Panchal 5 Reputation points Microsoft External Staff Moderator

2025-08-28T13:47:09.6033333+00:00
Hi Anthony Perkins,

Thanks for your question! Scaling delays in Synapse Dedicated SQL Pools can result from session cleanup, compute reconfiguration, or active workloads. To improve scaling speed:

Drain workloads before scaling.

Scale during off-peak hours.

Use Azure CLI or PowerShell to monitor status.

Avoid overlapping operations.

Also, consider reviewing your distribution strategy and resource class usage. If you're scaling from DW300 to DW1000 and have active jobs (e.g., on-prem data copies), that could explain the delay.

If you have any further questions, feel free to respond in this same thread.
Anthony Perkins 0 Reputation points

2025-08-28T15:38:34.5466667+00:00

Thank you for your help. Even after ensuring there is no activity, it still took over 45 minutes to scale back from DWU1000 to DW300.

The goal in scaling up was to see if we could resolve the errors we are seeing such as:

Msg 22225, Level 16, State 0, Line 1 An internal error (1206, 254) occurred. Please retry the operation. If the problem persists contact Microsoft Azure Customer Support. Operation cancelled by user. The statement has been terminated.

And:

Error: PdwManagedToNativeInteropException ErrorNumber: 46724, MajorCode: 467, MinorCode: 24, Severity: 20, State: 2, Exception of type 'Microsoft.SqlServer.DataWarehouse.Tds.PdwManagedToNativeInteropException' was thrown.

But it made no difference, and we are unable to re-index a very large table (1.2TB)

Obviously, we have a more serious problem than can be addressed here, so please go ahead and close this thread.

What do we need to do in order to get paid support from Microsoft, as currently when going through the portal we get messages like this:

Your cloud service provider manages your subscription. Contact your provider to open a support request.

And:

Your subscription has a technical support plan that can be used to create support requests related to Azure Active Directory, Information Protection, Multi-Factor Authentication and Intune only

Share via

Scaling up Synapse Dedicated Pools taking a long time

1 answer

Your answer