recurring simulation failures in ClogRemoteTLog.toml

[gxglass]

Saving off some notes as I could not see a very obvious clean way to fix this in limited time spent.

I believe this may reflect long standing brokenness. I dug up the below output from an internal CI report that runs nightly simulations. The repro args here might be useful. Mainly my intent is to note that we have been seeing failures on ClogRemoteTLog.toml on release-7.4 and probably on main for a long time.

Seen in foundationdb-nightly-clang-rhel9 for release-7.4 on Linux RHEL 9 on 2025-10-20

Result of foundationdb-nightly-clang-rhel9 for release-7.4 on Linux RHEL 9
• CodeBuild project: foundationdb-nightly-clang-rhel9
• Branch: release-7.4
• Commit ID: ef0e70739fbc29747454b1682b5b8694d61c4b20
• Duration: 7:33:43
• Result: :white_check_mark: SUCCEEDED
• Error: N/A
• Build Log terminal output (available for 30 days)
• Build Workspace zip file of the working directory (available for 30 days)
Joshua (Correctness) Results:
|   Passed |   Failed |
|----------|----------|
|   249997 |        3 |
:x: Failure Details:
| Test File                                           |       Seed |   Buggify |
|-----------------------------------------------------|------------|-----------|
...
| rare/ClogRemoteTLog.toml                            | 3581317426 |         1 |

clogremotetlog-quietdb-failure.md

[hiSandog]

Could the retained artifact include the exact simulation invocation, seed, buggify mode, and commit in a machine-copyable block? Since the referenced build logs expire after 30 days, preserving those inputs plus the first relevant trace event would keep the failure reproducible after the CI artifacts disappear.