diff options
| author | Oran Agra <oran@redislabs.com> | 2023-04-05 14:45:42 +0300 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-04-05 14:45:42 +0300 |
| commit | f263b6daf3a2672acc383dc34ed1ff1fe19da5a7 (patch) | |
| tree | ac50453d7a46470de1f5dbb82f1b4cb388030790 /.github | |
| parent | ccc86a91b7bb4ef84898dadfc9741ecb3aa8eadb (diff) | |
| download | redis-f263b6daf3a2672acc383dc34ed1ff1fe19da5a7.tar.gz | |
Increase threshold for flaky cache reclaim test (#12004)
This test produces 1GB of data and moves it around, and was expecting less
than 500kb to be present in the system page cache.
It sometimes fails with up to some 6mb in the page cache (0 in the actual RDB files),
increasing the threshold. It looks like some background tasks in the container are
occupying the page cache.
It is safe to ignore the above since we also explicitly check the pages of our dump.rdb
are not cached (matching `vmtouch -v` to `0%`).
An additional fix is to match ` 0%` (add space), so that we don't successfully match `10%`.
details in https://github.com/redis/redis/pull/11818
Diffstat (limited to '.github')
| -rw-r--r-- | .github/workflows/daily.yml | 14 |
1 files changed, 7 insertions, 7 deletions
diff --git a/.github/workflows/daily.yml b/.github/workflows/daily.yml index 266834675..5b7a86867 100644 --- a/.github/workflows/daily.yml +++ b/.github/workflows/daily.yml @@ -323,10 +323,10 @@ jobs: ./src/redis-cli -p 8080 save > /dev/null VMOUT=$(vmtouch -v /tmp/master/dump.rdb) echo $VMOUT - grep -q "0%" <<< $VMOUT + grep -q " 0%" <<< $VMOUT CACHE=$(grep -w file /sys/fs/cgroup/memory.stat | awk '{print $2}') echo "$CACHE" - if [ "$(( $CACHE-$CACHE0 ))" -gt "500000" ]; then exit 1; fi + if [ "$(( $CACHE-$CACHE0 ))" -gt "8000000" ]; then exit 1; fi echo "test replication doesn't increase cache" ./src/redis-cli -p 8081 REPLICAOF 127.0.0.1 8080 > /dev/null @@ -334,13 +334,13 @@ jobs: sleep 1 # wait for the completion of cache reclaim bio VMOUT=$(vmtouch -v /tmp/master/dump.rdb) echo $VMOUT - grep -q "0%" <<< $VMOUT + grep -q " 0%" <<< $VMOUT VMOUT=$(vmtouch -v /tmp/slave/dump.rdb) echo $VMOUT - grep -q "0%" <<< $VMOUT + grep -q " 0%" <<< $VMOUT CACHE=$(grep -w file /sys/fs/cgroup/memory.stat | awk '{print $2}') echo "$CACHE" - if [ "$(( $CACHE-$CACHE0 ))" -gt "500000" ]; then exit 1; fi + if [ "$(( $CACHE-$CACHE0 ))" -gt "8000000" ]; then exit 1; fi echo "test reboot doesn't increase cache" PID=$(cat /tmp/master/redis.pid) @@ -351,10 +351,10 @@ jobs: sleep 1 # wait for the completion of cache reclaim bio VMOUT=$(vmtouch -v /tmp/master/dump.rdb) echo $VMOUT - grep -q "0%" <<< $VMOUT + grep -q " 0%" <<< $VMOUT CACHE=$(grep -w file /sys/fs/cgroup/memory.stat | awk '{print $2}') echo "$CACHE" - if [ "$(( $CACHE-$CACHE0 ))" -gt "500000" ]; then exit 1; fi + if [ "$(( $CACHE-$CACHE0 ))" -gt "8000000" ]; then exit 1; fi test-valgrind-test: runs-on: ubuntu-latest |