Add setup script for elasticsearch, influxdb, redis

[czheo]

Add setup script for elasticsearch, influxdb, redis.
works on Linux only.

[at15]

why this is needed? does it require sudo?

[czheo]

yes, sudo is required.
elasticsearch checks max map count. without this, it won't boot up correctly.
https://www.elastic.co/guide/en/elasticsearch/reference/5.0/_maximum_map_count_check.html

[at15]

The ideal situation is check max_map_count size before ask sudo to change it just cat /proc/sys/vm/max_map_count will give you the value. though some transformation need to be made in shell. Anyway since *nix won't ask pwd for sudo in same shell, it's not a big deal.

[at15]

specify the version, same as influx and redis, though using latest is not that bad, but I don't want fetch the newest one, a working (most commonly used) one is enough

[czheo]

fixed.

[at15]

tested locally

[at15]

oops, some small problems @czheo

• update the readme file in script folder
• are we going to write all the time series logic for redis in our benchmark tool? if so how? following sentry or the redislab tech guide?

Thx

[czheo]

are we going to write all the time series logic for redis in our benchmark tool?

yes. we have to

if so how? following sentry or the redislab tech guide?

depends on what operations you'd like to cover.
Now, we only cover "count" and "moving avg".
count is native in redis.
we need find some sample to implement moving avg. any sample is fine to make a comparison.
Such as:
https://redislabs.com/ebook/redis-in-action/part-2-core-concepts-2/chapter-5-using-redis-for-application-support/5-2-counters-and-statistics/5-2-2-storing-statistics-in-redis
https://www.redisgreen.net/library/runningavg.html

[at15]

@czheo I think sentry have some existing logic https://github.com/getsentry/sentry/tree/master/src/sentry/tsdb

[czheo]

@at15 didn't find moving avg in their code