mirror of https://github.com/valkey-io/valkey
61b63ab989
This extends the `__rand_int__` pattern to add `__rand_1st__` through
`__rand_9th__`.
For the new placeholders, multiple occurrences within a command will
have the same value.
With the `--sequential` option, each placeholder will use separate
counters.
For `__rand_int__`, the behavior is unchanged - multiple occurrences of
the pattern within a command will have different values.
Examples of using the `__rand_int__` with `--sequential`:
```
$ ./valkey-benchmark -r 300 -n 300 --sequential -q -- set key__rand_int__ val__rand_int__
set key__rand_int__ val__rand_int__: 60000.00 requests per second, p50=0.391 msec
$ ./valkey-cli info keyspace
# Keyspace
db0:keys=150,expires=0,avg_ttl=0
$ ./valkey-cli get key000000000050
"val000000000051"
```
For the new patterns, multiple occurrences within the command will have
the same value.
```
$ ./valkey-benchmark -r 300 -n 300 --sequential -q -- set key__rand_1st__ val__rand_1st__
set key__rand_int__ val__rand_int1_: 60000.00 requests per second, p50=0.383 msec
$ ./valkey-cli info keyspace
# Keyspace
db0:keys=300,expires=0,avg_ttl=0
$ ./valkey-cli get key000000000050
"val000000000050"
```
I also fixed the zadd benchmark so it produces the expected number of
keys when used with `--sequential`. (By using the same counter twice per
command it was effectively counting by twos)
I made this for myself but raised a PR in case y'all like it. 🙂
---------
Signed-off-by: Rain Valentine <rsg000@gmail.com>
|
||
|---|---|---|
| .. | ||
| format.yml | ||
| typos.toml | ||