You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@zookeeper.apache.org by gp <bi...@163.com> on 2017/07/25 05:35:31 UTC
Is there a benchmark performance test to reveal how the disk's iops
effect zookeeper's tps/qps?
As the document says
"incorrect placement of transasction log
The most performance critical part of ZooKeeper is the transaction log. ZooKeeper syncs transactions to media before it returns a response. A dedicated transaction log device is key to consistent good performance. Putting the log on a busy device will adversely effect performance..."
However, is there a benchmark perf test to reveal how the disk's iops effect zookeeper's tps/qps?
Re: Is there a benchmark performance test to reveal how the disk's
iops effect zookeeper's tps/qps?
Posted by Michael Han <ha...@cloudera.com>.
For public benchmark:
* ZK has a systest in src/java/systest.
* Checkout https://coreos.com/blog/performance-of-etcd.html. There is a
github link on the benchmark.
If you end up writing your own benchmark, please consider contribute it
back to open source :)
On Mon, Jul 24, 2017 at 10:35 PM, gp <bi...@163.com> wrote:
> As the document says
> "incorrect placement of transasction log
> The most performance critical part of ZooKeeper is the transaction log.
> ZooKeeper syncs transactions to media before it returns a response. A
> dedicated transaction log device is key to consistent good performance.
> Putting the log on a busy device will adversely effect performance..."
> However, is there a benchmark perf test to reveal how the disk's iops
> effect zookeeper's tps/qps?
>
>
>
>
>
>
>
>
>
>
--
Cheers
Michael.