summaryrefslogtreecommitdiff
path: root/scripts/benchmark
diff options
context:
space:
mode:
authordrebs <drebs@leap.se>2017-07-05 11:43:58 -0300
committerKali Kaneko <kali@leap.se>2017-07-07 21:25:21 +0200
commitda7745639bb7d38e25c7c0f0e1e405cb4c3f4c7e (patch)
treeb83bad3dc37330e1cc61c84e91f0d99474587e25 /scripts/benchmark
parent2f190ac7e0dfbf5e15e4e28642d9106e835eee31 (diff)
[benchmarks] add README to benchmark-all-commits
Diffstat (limited to 'scripts/benchmark')
-rw-r--r--scripts/benchmark/README.md31
1 files changed, 31 insertions, 0 deletions
diff --git a/scripts/benchmark/README.md b/scripts/benchmark/README.md
new file mode 100644
index 00000000..0910219a
--- /dev/null
+++ b/scripts/benchmark/README.md
@@ -0,0 +1,31 @@
+Benchmark all commits
+=====================
+
+The script in this directory is used to run the benchmarking tests for all
+commits in the repository.
+
+It is actually not possible to run all benchmarking tests for all commits
+because:
+
+ - at some points in the repo history, the tested code didn't exist.
+
+ - at some poitns in the repo history, the teste code might have been
+ refactored and the api would not match the one used in tests.
+
+So what this script does is the following:
+
+ - check out .gitlab-ci.yml from current master.
+ - use gitlab-runner to exec a build locally.
+ - store timestamp and exit status of each run.
+ - checkout previous commit and iterate.
+
+How to use the script
+---------------------
+
+Once you have gitlab-runner setup in the machine and the elasticsearch
+credentials, use the following commands to run the script:
+
+ git clone https://0xacab.org/leap/soledad
+ cd soledad/
+ export PYTEST_OPTS="--benchmark-autosave --benchmark-storage=\"elasticsearch+https://USER:PASSWORD@moose.leap.se:9200\""
+ scripts/benchmark/benchmark-all-commits.sh