# Benchmark using Sysbench

**URL:** <https://forum.yugabyte.com/t/benchmark-using-sysbench/739>\
**Category:** General\
**Created:** [June 23, 2020, 8:37am UTC](https://forum.yugabyte.com/t/benchmark-using-sysbench/739 "2020-06-23T08:37:25Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Quy\_n\_Ph\_m](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.yugabyte.com/quy_n_ph_m/32/227_2.png) [@Quy\_n\_Ph\_m](https://forum.yugabyte.com/u/Quy_n_Ph_m)\
**Post date:** [June 23, 2020, 8:37am UTC](https://forum.yugabyte.com/t/benchmark-using-sysbench/739/1 "2020-06-23T08:37:25Z")

</div>

I am new with YugabyteDB and I have followed the instructions for benchmarking using sysbench here:

> **[Benchmark YSQL performance using sysbench](https://docs.yugabyte.com/preview/benchmark/sysbench-ysql/)**
>
> Benchmark YSQL performance using sysbench.

However, I am confused about the **pgsql-host** and **pgsql-port** in the command. How can I get those values after deploying YB-Tservers and YB-Master? Is it the host and port of one of YB-Tserver nodes or YB-Master nodes?

Many thanks,

---

<div class="post-metadata">

**Author:** ![dorian\_yugabyte](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.yugabyte.com/dorian_yugabyte/32/206_2.png) [@dorian\_yugabyte](https://forum.yugabyte.com/u/dorian_yugabyte)\
**Post date:** [June 23, 2020, 9:18am UTC](https://forum.yugabyte.com/t/benchmark-using-sysbench/739/2 "2020-06-23T09:18:43Z")

</div>

Hi @Quy_n_Ph_m

Welcome to YugabyteDB Forum!

`--pgsql-host` is the ip of the `yb-tserver` process. Clients always connect to `yb-tservers`.  
`--pgsql-port` is the port of YSQL, which by default is `5433`. It’s the same address+port as set in [–pgsql\_proxy\_bind\_address](https://docs.yugabyte.com/latest/reference/configuration/yb-tserver/#pgsql-proxy-bind-address) in yb-tserver.

Regards,  
Dorian  
Technical Support Engineer

---

<div class="post-metadata">

**Author:** ![Quy\_n\_Ph\_m](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.yugabyte.com/quy_n_ph_m/32/227_2.png) [@Quy\_n\_Ph\_m](https://forum.yugabyte.com/u/Quy_n_Ph_m)\
**Post date:** [June 30, 2020, 7:21am UTC](https://forum.yugabyte.com/t/benchmark-using-sysbench/739/3 "2020-06-30T07:21:41Z")

</div>

Hi,  
My deployment model 3 YB-TServers and 3 YB-Masters distributed on 3 nodes. Also we have 1 HAProxy on another node to perform load balance between YB-TServers so clients call requests to HAProxy instead of YB-TServers. Because we use sysbench to benchmark which takes only one host and port in the config file so we use the host and port of HAProxy. Does that model degrade YugabyteDB performance?

---

<div class="post-metadata">

**Author:** ![dorian\_yugabyte](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.yugabyte.com/dorian_yugabyte/32/206_2.png) [@dorian\_yugabyte](https://forum.yugabyte.com/u/dorian_yugabyte)\
**Post date:** [July 3, 2020, 5:05pm UTC](https://forum.yugabyte.com/t/benchmark-using-sysbench/739/4 "2020-07-03T17:05:19Z")

</div>

@Quy_n_Ph_m just like I explained in [Should we use HAProxy to load balance between YB-TServers? · Issue #4923 · yugabyte/yugabyte-db · GitHub](https://github.com/yugabyte/yugabyte-db/issues/4923), it adds latency because of more network hops, but the databases is not slower. The throughput should be the same.
