diff options
| author | Alex Auvolat <alex@adnab.me> | 2023-10-20 15:00:10 +0200 |
|---|---|---|
| committer | Alex Auvolat <alex@adnab.me> | 2023-10-20 15:00:10 +0200 |
| commit | f5b09727815523a1bd4ba5f62d892b2b45b5bed6 (patch) | |
| tree | 6ba6492e2df9663007f6af09f8108532a284ded5 /script/jepsen.garage/README.md | |
| parent | d148b83d4f440dc79b2ed08eaa171aca0e2037b0 (diff) | |
| download | garage-f5b09727815523a1bd4ba5f62d892b2b45b5bed6.tar.gz garage-f5b09727815523a1bd4ba5f62d892b2b45b5bed6.zip | |
jepsen: register crdt read-after-write is fixed with deleteobject patch
Diffstat (limited to 'script/jepsen.garage/README.md')
| -rw-r--r-- | script/jepsen.garage/README.md | 11 |
1 files changed, 5 insertions, 6 deletions
diff --git a/script/jepsen.garage/README.md b/script/jepsen.garage/README.md index da6f0b77..4c3c70b3 100644 --- a/script/jepsen.garage/README.md +++ b/script/jepsen.garage/README.md @@ -45,7 +45,7 @@ clocks are scrambled. ### Register linear, with timestamp patch -Command: `lein run test --nodes-file nodes.vagrant --time-limit 60 --rate 20 --concurrency 20 --workload reg1 --ops-per-key 100 -I` +Command: `lein run test --nodes-file nodes.vagrant --time-limit 60 --rate 20 --concurrency 20 --workload reg1 --ops-per-key 100 --patch tsfix1` Results: @@ -62,15 +62,13 @@ Results: fails with a simple clock-scramble nemesis. Explanation: old values are not overwritten correctly when their timestamps are in the future. -### Read-after-write CRDT register model, with timestamp patch +### Read-after-write CRDT register model, with timestamp patch (v2 with DeleteObject fix as well) -Command: `lein run test --nodes-file nodes.vagrant --time-limit 60 --rate 100 --concurrency 100 --workload reg2 --ops-per-key 100 -I` +Command: `lein run test --nodes-file nodes.vagrant --time-limit 60 --rate 100 --concurrency 100 --workload reg2 --ops-per-key 100 --patch tsfix2` Results: -- Failures with clock-scramble nemesis + partition nemesis ???? TODO INVESTIGATE - -> the issue seems to be only after DeleteObject (deletions are not always taken into account), - the issue does not appear if we are using only PutObject with an actual object content +- No failures with clock-scramble nemesis + partition nemesis - TODO: layout reconfiguration nemesis @@ -123,6 +121,7 @@ The inconsistencies seemed to always happenned after writing a nil value, which instead of a PutObject. By removing the possibility of writing nil values, therefore only doing PutObject calls, the issue disappears. There is therefore an issue to fix in DeleteObject. +The issue in DeleteObject seems to have been fixed by commit `c82d91c6bccf307186332b6c5c6fc0b128b1b2b1` ## License |