PMM-15298: RTA elapsed time filter accepts numbers only

[mattiasimonato]

https://jira.percona.com/browse/PMM-15298

• PMM-15298: RTA elapsed time filter accepts numbers only percona/pmm#5955

[JNKPercona]

Server docker: perconalab/pmm-server-fb:PR-4567-8172997
Client docker: perconalab/pmm-client-fb:PR-4567-8172997
Client tarball: https://s3.us-east-2.amazonaws.com/pmm-build-cache/PR-BUILDS/pmm-client/pmm-client-PR-4567-8172997.tar.gz
Staging instance: https://pmm.cd.percona.com/job/pmm3-aws-staging-start/parambuild/?DOCKER_VERSION=perconalab/pmm-server-fb:PR-4567-8172997&CLIENT_VERSION=https://s3.us-east-2.amazonaws.com/pmm-build-cache/PR-BUILDS/pmm-client/pmm-client-PR-4567-8172997.tar.gz

[JNKPercona]

API tests have succeded: https://pmm.cd.percona.com/job/pmm3-api-tests/7377/

[travagliad]

The three red @pmm-ps-integration tests (PMM-T2028 ×2, PMM-T1897) reproduced on a throwaway VM and are not caused by your change — they are pmm-qa test bugs. pmm-qa main commit 6f57cd2 (#1433) moved pmm-framework's default PS version from 8.0 to 8.4 a few hours before your FB build ran; the same 10 tests passed on main at 02:22 UTC with identical setup args and PS 8.0.

MySQL 8.4 renamed the SHOW SLAVE STATUS columns (so the metrics are mysql_slave_status_replica_*, not ..._slave_*), and the 8.4 mysql client preserves comments by default, so the QAN seed file's comment lines land in the slow log and inflate the query count from 19 to 20.

Fix PR: percona/pmm-qa#1436. Nothing is needed from you — once that merges, a rebuild of this FB should come back clean.

---

Generated by Claude Code