cdist performance 2.0.0-rc4

Signed-off-by: Nico Schottelius <nico@kr.ethz.ch>
This commit is contained in:
Nico Schottelius 2011-09-14 01:03:04 +02:00
parent 56520a3807
commit 488bd9854f
2 changed files with 64 additions and 0 deletions

View file

@ -0,0 +1,64 @@
[[!meta title="Performance of cdist 2.0.0-rc4"]]
As you may have notice, [[cdist|software/cdist]] 2.0.0
has been rewritten in Python due to performance issues with
fork(). This article shows the first performance results of
the new implementation
## Background
All configurations were done using the productive
configuration of the Systems Group in the ETH Zurich.
All hosts were configured from a cable modem connection
on my Lenovo X201, with 2 cores, 4 threads
(Intel(R) Core(TM) i5 CPU M 520 @ 2.40GHz) and 4 GiB
RAM. The configuration contains 189 objects, of which
142 are of type **__package**.
The duration was taken from cdist output and the measurements
were taken by deploying to one, two, ..., thirty one hosts in
parallel:
(
set --;
for host in ikq02.ethz.ch ikq03.ethz.ch ikq04.ethz.ch ikq05.ethz.ch \
ikq06.ethz.ch ikq07.ethz.ch ikr01.ethz.ch ikr02.ethz.ch ikr03.ethz.ch \
ikr05.ethz.ch ikr07.ethz.ch ikr09.ethz.ch ikr10.ethz.ch ikr11.ethz.ch \
ikr13.ethz.ch ikr14.ethz.ch ikr15.ethz.ch ikr16.ethz.ch ikr17.ethz.ch \
ikr19.ethz.ch ikr20.ethz.ch ikr21.ethz.ch ikr23.ethz.ch ikr24.ethz.ch \
ikr25.ethz.ch ikr26.ethz.ch ikr27.ethz.ch ikr28.ethz.ch ikr29.ethz.ch \
ikr30.ethz.ch ikr31.ethz.ch;
do
set -- "$@" $host;
cdist -c ~/p/cdist-nutzung -p "$@";
done
) 2>&1 | tee ~/benchmark-home
## The results
Deploying to one host took 72 seconds, to three hosts 77 seconds,
to 6 hosts 108 seconds and to 31 hosts 282 seconds, as can be seen
in the following graph:
[[!img cdist-2.0.0-rc4-graph.png size="400x300" alt="Cdist graph"]]
In a sequential run, it would have taken 2232 seconds to deploy
to 31 hosts. At higher parallel configurations (>10), it can be seen
that cdist becomes CPU bound.
Although deploying to 31 hosts takes much longer than 1 host, it is
still much faster than the linear case.
## Conclusion
* 72 seconds is still pretty long for one run, but can easily be improved
by comparing the configuration against the cache and only run the new
parts (or nothing if rerunning)
* Cdist can deploy to more than 30 hosts on 2 cores/4 threads within 5 minutes
* As cdist runs massively parallel, it can utilise all existing CPUs
* And even more important: cdist can scale out, if you add more CPUs (or machines),
it can simply use them to be able to deploy to more machines in parallel
[[!tag config eth unix]]

Binary file not shown.

After

Width:  |  Height:  |  Size: 24 KiB