sqitchers / sqitch

100%

LAST BUILD ON BRANCH develop

branch: SELECT

CHANGE BRANCH

x

• No branch selected
• adjust-workflow-triggers
• badges
• bundle
• clickhouse
• clickhouse-cli
• clickhouse-config
• coverage
• cron-workflow
• develop
• docker
• engine-mock-config
• firebird
• fix-cover
• fixes
• free-cover
• issue-828
• issue-875
• just-clickhouse-binary
• main
• master
• more-coverage
• mysql-registry-uri
• option-unification
• pgsql-versions
• postgres-collate
• prep-v1.6.0
• recoverage
• release-prep
• snowflake
• snowflake-keypair
• sqlite-coverage
• target-variables
• testconfig
• travis-update
• tutorial
• update-libintl
• v0.9998
• yugabyte-lock

Build # 21339916894

Build Type

push

github

Committed by theory

Commit Message

Start ClickHouse after deleting files

The coverage workflow started complaining of a shortness of disk space
before it got to clearing the disk space. The likely culprit is the
services, which all start before the steps have a chance to clear space.

So move starting the ClickHouse server in coverage workflow from
`services` to `steps`. This requires that it always map ports 8123 and
9000, rather than let the runner find ports, but it seems to work, and
has worked for CockroachDB, which follows the same pattern for some
time.

While at it, remove the `-t` option from the call to
`apt.postgresql.org.sh` in the Postgres installation script, as it fails
to work properly with Postgres 18.

Add "Installing X" output to each of the install scripts, to make it
easier, in the future, to see which ones have run in the "Setup Clients"
step of the coverage workflow.

Fix the SQLite version string for 2026, now that there is a first
release.

Run Details

4828 of 4828 relevant lines covered (100.0%)

70.63 hits per line