Alexander Bayandin
|
a0b3990411
|
Retry data ingestion scripts on connection errors (#4382)
## Problem
From time to time, we're catching a race condition when trying to upload
perf or regression test results.
Ref:
- https://neondb.slack.com/archives/C03H1K0PGKH/p1685462717870759
- https://github.com/neondatabase/cloud/issues/3686
## Summary of changes
Wrap `psycopg2.connect` method with `@backoff.on_exception`
contextmanager
|
2023-06-13 22:33:42 +01:00 |
|
Alexander Bayandin
|
35bb10757d
|
scripts/ingest_perf_test_result.py: increase connection timeout (#4329)
## Problem
Sometimes default connection timeout is not enough to connect to the DB
with perf test results, [an
example](https://github.com/neondatabase/neon/actions/runs/5064263522/jobs/9091692868#step:10:332).
Similar changes were made for similar scripts:
- For `scripts/flaky_tests.py` in
https://github.com/neondatabase/neon/pull/4096
- For `scripts/ingest_regress_test_result.py` in
https://github.com/neondatabase/neon/pull/2367 (from the very
begginning)
## Summary of changes
- Connection timeout increased to 30s for
`scripts/ingest_perf_test_result.py`
|
2023-05-24 10:11:24 -04:00 |
|
Alexander Bayandin
|
83dca73f85
|
Store Allure tests statistics in database (#2367)
|
2022-09-07 14:16:48 +01:00 |
|
Alexander Bayandin
|
4c2bb43775
|
Reformat all python files by black & isort
|
2022-08-22 14:57:09 +01:00 |
|
Dmitry Rodionov
|
ccf3c8cc30
|
store performance test results in our staging cluster to be able to
visualize them in grafana
|
2022-02-10 13:12:19 +03:00 |
|