Don't miss your chance to take exam DP-600 or DP-700 on us!
Request nowFabric Data Days Monthly is back. Join us on March 26th for two expert-led sessions on 1) Getting Started with Fabric IQ and 2) Mapping & Spacial Analytics in Fabric. Register now
When trying to run a realistic production query on Fabric Warehouse that we currently run on SQL Server I get
Msg 15802, Level 16, State 16, Line 1
Request to perform an external distributed computation has failed with error "Assertion failed: Invalid child output DOP".
This was the first query we tried..
I have done some playing, if i remove any construct in the query using COUNT(DISTINCT ...) there is no failure.
@JoshAsh could you open a support case for this please? We'd like to look deeper into this.
Hey Kevin
Done - 2306200030005070
Hi Josh,
The queries from SQL Server would not execute as is on Synapse WH as the TSQL has some changes.
do you have a sample of the script you are trying to execute?
I've had to change column and table names for security reasons.
SELECT
COUNT(DISTINCT CASE WHEN UPPER(fee.blah) LIKE '%AMEP%' THEN c.blah ELSE NULL END) as DISTINCT_BLAH_COUNT,
MAX(c.blah) as blah
FROM dbo.TABLE1 a
INNER JOIN dbo.TABLE2 d ON a.blah = d.blah
INNER JOIN dbo.TABLE3 b ON a.blah = b.blah
INNER JOIN dbo.TABLE4 dt ON a.blah = dt.blah
INNER JOIN dbo.TABLE5 att ON a.blah = att.blah
INNER JOIN dbo.TABLE6 c ON a.blah = c.blah
INNER JOIN dbo.TABLE7 fc ON a.blah = fc.blah
INNER JOIN dbo.TABLE8 age ON a.blah = age.blah
LEFT JOIN dbo.TABLE9 ft ON a.blah = ft.blah
LEFT JOIN dbo.TABLE10 fee ON a.blah = fee.blah
LEFT JOIN dbo.TABLE11 ccs ON a.blah = ccs.blah
WHERE
d.blah in ('Centre','Discontinued')
AND dt.blah >= 2019
GROUP BY
a.blah,
c.blah,
fc.blah,
dt.blah,
dt.blah
But this is the simplest version I can make that is erroring. It's not really a complex query as you can see
Another errror happening occasionally is
Request to perform an external distributed computation has failed with error "Assertion failed: Degree of partitioned parallelism can only be 1 when the child operator distribution is full or when the session dop is 1".
This is happening due to the COUNT(DISTINCT xyz) function as the data is distributed on Synapse Warehouse.
Can you try using APPROX_COUNT_DISTINCT instead and let me know if that works?
APPROX_COUNT_DISTINCT (Transact-SQL) - SQL Server | Microsoft Learn
Share feedback directly with Fabric product managers, participate in targeted research studies and influence the Fabric roadmap.
Check out the February 2026 Fabric update to learn about new features.
| User | Count |
|---|---|
| 2 | |
| 2 | |
| 1 | |
| 1 | |
| 1 |