Benchmark utils for Redis Cluster
Switch branches/tags
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
spec
src
.gitignore
.travis.yml
LICENSE
Makefile
README.md
bench.toml
shard.yml

README.md

redis-cluster-benchmark Build Status

Benchmark utils for Redis Cluster that is inspired by redis-benchmark.

Installation

Static Binary is ready for x86 linux

Compile from source

  • crystal-0.26.1
% shards update  # first time only
% make
% cp bin/redis-cluster-benchmark /usr/local/bin/

Config

  • See: bench.toml
[redis]
clusters = "127.0.0.1:7001,127.0.0.1:7002"  # at least one node
# password = "secret"

[bench]
requests = 100000  # number of requests for each commands
keyspace = 10000   # max number of key for `__rand_int__`
tests    = "set __rand_int__ __rand_int__"
after    = "set result/last __result__"
# after    = "rpush results __result__"
qps      = 5000
# debug = true

[report]
interval_sec = 1
verbose = true

Run

% redis-cluster-benchmark bench.toml
=== set __rand_int__ __rand_int__ ===
17:19:52 [5.0%] 5002/100000 (4989.3 qps)
17:19:53 [10.0%] 10003/100000 (4996.1 qps)
...
17:20:11 [95.1%] 95086/100000 (5000.9 qps)
17:20:12 done 100000 in 20.8 sec (0.0)
SET: 5820.68 rps (OK: 100000, KO: 0)
  • Algthough we sent 100,000 requests, only 10,000 keys are created due to keyspace = 10000.
% rcm -u :7001 info count
893550 [127.0.0.1:7001]  cnt(3334)
bd4456 [127.0.0.1:7002]  cnt(3330)
87b981 [127.0.0.1:7003]  cnt(3336)

After Hook

  • We can set after hook by after field in bench section.
  • In above example, benchmark result has been set in redis via special keyword __result__.
% redis-cli -c -p 7001 get result/last
"1977.6 qps (OK:100000, KO:0) [02:53:26 +51s]"
  • Of course, apending logs is good idea by rpush
after    = "rpush results __result__"

Error Reporting

12:00:26 [94.6%] 946044/1000000 (4993.2 qps) # KO: 109
12:00:27 [95.1%] 951045/1000000 (4992.1 qps) # KO: 270
12:00:28 [95.6%] 956046/1000000 (4991.9 qps) # KO: 110
12:00:29 [96.1%] 961047/1000000 (4994.8 qps)
12:00:30 [96.6%] 966048/1000000 (4994.4 qps)
12:00:31 [97.1%] 971049/1000000 (4990.4 qps)
12:00:32 [97.6%] 976050/1000000 (4992.0 qps)
12:00:33 [98.1%] 981051/1000000 (4991.3 qps)
12:00:34 [98.6%] 986052/1000000 (4997.0 qps)
12:00:35 [99.1%] 991053/1000000 (4995.3 qps)
12:00:36 [99.6%] 996054/1000000 (4993.2 qps)
12:00:36 done 1000000 in 201.9 sec (4953.0 qps) # KO: 818
SET: 4953.0 qps (OK:999182, KO:818) [11:57:15 +3m21s] # RedisError: OOM command not allowed when used
 memory > 'maxmemory'.
 (ERRORS)
   RedisError: OOM command not allowed when used memory > 'maxmemory'.
   === [AFTER] RPUSH results __result__ ===

Roadmap

0.4

  • Reconnect after errors
  • Multi Clients

0.5

  • Connection Setting (use keep alive or not)
  • Protocol Setting (use pipeline or not)

Contributing

  1. Fork it ( https://github.com/maiha/redis-cluster-benchmark/fork )
  2. Create your feature branch (git checkout -b my-new-feature)
  3. Commit your changes (git commit -am 'Add some feature')
  4. Push to the branch (git push origin my-new-feature)
  5. Create a new Pull Request

Contributors

  • maiha maiha - creator, maintainer