SELECT 💣();

[bayandin]

Problem

We want to be able to test how our infrastructure reacts on segfauls in Postgres (for example, we collect core dumps, get some required logs/metrics, etc)

Summary of changes

• Add trigger_segfauls function to neon_test_utils to trigger a segfault in Postgres
• Add trigger_panic function to neon_test_utils to trigger SIGABRT (by using `elog(PANIC, ...))
• Fix cleanup logic in regression tests in endpoint crashed

Checklist before requesting a review

• I have performed a self-review of my code.
• If it is a core feature, I have added thorough tests.
• Do we need to implement analytics? if so did you add the relevant metrics to the dashboard?
• If this PR requires public announcement, mark it with /release-notes label and add several sentences in this section.

Checklist before merging

• Do not forget to reformat commit message to not include the above checklist

[github-actions]

3042 tests run: 2927 passed, 0 failed, 115 skipped (full report)

---

Flaky tests (2)

Postgres 16

• test_sharding_split_compaction[None]: debug

Postgres 15

• test_peer_recovery: debug

Code coverage* (full report)

• functions: 32.6% (6929 of 21274 functions)
• lines: 50.0% (54437 of 108936 lines)

* collected from Rust tests only

---

_{The comment gets automatically updated with the latest test results
ce27f53 at 2024-07-05T11:42:10.821Z :recycle:}

[kelvich]

awesome!

[kelvich]

And you, perhaps, need to bump extension version and write migration script.

[MMeent]

Why do we need this? I don't see it.

[kelvich]

Why do we need this? I don't see it.

version bump or function to segfault? If latter then to test crash behavior (e.g. will compute_ctl restart postgres) and test coredump collection / printing

[bayandin]

Why do we need this? I don't see it.

Added it to the description:

We want to be able to test how our infrastructure reacts on segfauls in Postgres (for example, we collect core dumps, get some required logs/metrics, etc)

Haven't noticed Stas' answer, but the idea is the same

[MMeent]

Could we then also have one that does elog(PANIC, "...") to simulate such types of bad situations?

[bayandin]

And you, perhaps, need to bump extension version and write migration script.

I bumped the version. Don't think the migration script is required for this one (we're pretty chill about it for test extension, for example, the previous changes 75c84c8)

[skyzh]

Note that this extension can be directly used by our users, so I would avoid putting things that might crash compute in that... There was an attempt to remove user access but autoscaling tests need this extension so it was reverted. #7679