In the Windows HLK 24H2 (updated May 2025), the test "Installable File System Filter Test" is failing on Windows Server 2025 24H2. The following is present in the .wtl file regarding the failure:
Error 6/24/2025 10:21:34.726 PM E40.1A1C : +TEST+SEV2 : Test :FlushBuffersRootTest
Group :SectionsCaching
Status :C000004E (IFSTEST_TEST_STATISTICS_ERROR)
Description :{Msg# seccache!flushrot!18} The first write statistic
was 1310720. The second 1310720. The total number of
bytes modified on the files is 291088. It does not
appear that all data was flushed.
WDKTUID: A72C4D26-977B-19CB-2E11-D6F93917D718
I've noted that with down-level HLK controllers & test systems, the same test completes on Windows Server 2022 & 2019 with the same underlying failure, yet there is a filter present which allows the failure to be suppressed and the status glyph has a circle around the checkmark.
I've ensured that the latest available filter update, downloaded 2025-06-25, has been applied to my controller, but the failure is not being filtered and converted to a successful result.
I'm not sure this is helpful, but I can say that I just ran these tests (literally...like right now) as part of certifying our own product and the IFS tests passed for all target file systems:
Test :FlushBuffersRootTest Group :SectionsCaching Status :00000000 (IFSTEST_SUCCESS) WDKTUID: A72C4D26-977B-19CB-2E11-D6F93917D718
HLK Controller is the same VHLK you're using.
Target is Server 2025 June update (10.0.26100.4349) Hyper-V VM.
Target drives are VHDs attached from the host:
-
One VHD with NTFS, CNTFS, FAT16, FAT32, exFAT, and UDFS volumes.
-
Two VHDs which make up a Storage Spaces mirror which in turn has an ReFS volume
Now, that being said, I did initially have problems getting the HLK to run any IFS tests at all. It just wouldn't recognize any filter as a File System Filter and wouldn't offer me the tests. This was a Server 2025 from March as the HLK Client, so I decided to just delete everything and start again from the beginning with the June ISO. I have no idea if this made a difference (and have no intention of going back and debugging further 
)
I didn't know there was a June update released for HLK 24H2. The issue may well have been fixed in it. I'll look at the HLK downloads and get it installed if I see it as being available. If you look at your test results which show success, is there any indication that they have been filtered to suppress an underlying error (that's expected to occur) just like I see on Windows Server 2019 1809 & Windows Server 22 21H2?
Oh, BTW, do you know where to get the older HLK installer to be able to run tests on Windows Server 2016 1803 systems? It's still on Microsoft's support matrix with LTSC support ending on 2027-01-12, but I don't see an HLK available to test it.
I double checked on the HLK downloads and don't see a specific option for a June update. There's no ISO image for it and the download link still shows May 2025 as the update. Did you simply download the HLKSetup.exe file and then find that the file it downloads are actually for a June 2025 update?
Never mind about the down-level HLK question. I had to do some digging in my MSDN subscription downloads and found an ISO image for the 1803 HLK to test with Windows Server 2016 1803.
No, sorry, my HLK Client was Server 2025 June update
The HLK Server was the May update.
Also, having my ReFS volume in a mirror was a bad idea because it causes the Scrub Tests to fail (they want a non-RAID disk). Once took that out of a mirror everything passed.