Updated S3 API Benchmark (markdown)

Chris Lu 2020-11-29 17:25:06 -08:00
parent d8f62791dc
commit bc7f5eb259

@ -44,6 +44,60 @@ warp: Cleanup Done.
```
# Another test with warp
Here is a user-provided `run-warp.sh` file.
```
#!/bin/bash -e
if [[ "$#" -lt 1 ]]
then
echo "Usage: $0 count [options]"
exit 1
fi
COUNT="$1"
shift
COUNTER=0
while [ $COUNTER -lt $COUNT ]; do
let COUNTER=COUNTER+1
warp client "127.0.0.1:576${COUNTER}" \
1> warp-client-"${COUNTER}".out \
2> warp-client-"${COUNTER}".err &
done
sleep 5
warp "$@" \
--host "localhost:8333" \
--access-key "some_access_key1" --secret-key "some_secret_key1" \
--bucket warp-"${COUNTER}-$(date +%s)" \
--warp-client=127.0.0.1:576{1...$COUNT} \
1> warp.out \
2> warp.err
```
Start the warp this way:
```
./run-warp.sh 4 list --concurrent 5 --objects 100000 --obj.size=4KiB --duration 60s
```
Here is the result:
```
warp: Benchmark data written to "warp-remote-2020-11-29[165701]-gsYs.csv.zst"
Operation: PUT
* Average: 15.35 MiB/s, 3929.68 obj/s
Throughput, split into 101 x 1s:
* Fastest: 17.1MiB/s, 4387.13 obj/s
* 50% Median: 15.8MiB/s, 4045.38 obj/s
* Slowest: 12.8MiB/s, 3280.82 obj/s
```
# Test with Hotsauce S3 Benchmark
Found this https://github.com/markhpc/hsbench tool which seems easy to use.