<div dir="ltr">If there was going to be a benchmark suite for keystone, I'd suggest contributing it to keystoneclient rather than keystone itself. That said, here's an example of a one liner exercising v3 auth:<div>
<br></div><div><font face="courier new, monospace">$ keystone --os-token=ADMIN --os-endpoint=<a href="http://localhost:35357/v2.0/">http://localhost:35357/v2.0/</a> bootstrap --user-name=admin --pass=admin --tenant-name=admin --role-name=admin<br>
</font></div><div><div><font face="courier new, monospace">$ cat auth-v3.json </font></div><div><font face="courier new, monospace">{"auth": {"identity": {"methods": ["password"], "password": {"user": {"name": "admin", "password": "admin", "domain": {"id": "default"}}}}, "scope": {"project": {"name": "admin", "domain": {"id": "default"}}}}}</font></div>
</div><div><div><font face="courier new, monospace">This is ApacheBench, Version 2.3 <$Revision: 1373084 $></font></div><div><font face="courier new, monospace">Copyright 1996 Adam Twiss, Zeus Technology Ltd, <a href="http://www.zeustech.net/">http://www.zeustech.net/</a></font></div>
<div><font face="courier new, monospace">Licensed to The Apache Software Foundation, <a href="http://www.apache.org/">http://www.apache.org/</a></font></div><div><font face="courier new, monospace"><br></font></div><div><font face="courier new, monospace">Benchmarking 127.0.0.1 (be patient).....done</font></div>
<div><font face="courier new, monospace"><br></font></div><div><font face="courier new, monospace"><br></font></div><div><font face="courier new, monospace">Server Software: </font></div><div><font face="courier new, monospace">Server Hostname: 127.0.0.1</font></div>
<div><font face="courier new, monospace">Server Port: 35357</font></div><div><font face="courier new, monospace"><br></font></div><div><font face="courier new, monospace">Document Path: /v3/auth/tokens</font></div>
<div><font face="courier new, monospace">Document Length: 463 bytes</font></div><div><font face="courier new, monospace"><br></font></div><div><font face="courier new, monospace">Concurrency Level: 16</font></div>
<div><font face="courier new, monospace">Time taken for tests: 16.192 seconds</font></div><div><font face="courier new, monospace">Complete requests: 100</font></div><div><font face="courier new, monospace">Failed requests: 0</font></div>
<div><font face="courier new, monospace">Write errors: 0</font></div><div><font face="courier new, monospace">Total transferred: 66700 bytes</font></div><div><font face="courier new, monospace">Total body sent: 36200</font></div>
<div><font face="courier new, monospace">HTML transferred: 46300 bytes</font></div><div><font face="courier new, monospace">Requests per second: 6.18 [#/sec] (mean)</font></div><div><font face="courier new, monospace">Time per request: 2590.670 [ms] (mean)</font></div>
<div><font face="courier new, monospace">Time per request: 161.917 [ms] (mean, across all concurrent requests)</font></div><div><font face="courier new, monospace">Transfer rate: 4.02 [Kbytes/sec] received</font></div>
<div><font face="courier new, monospace"> 2.18 kb/s sent</font></div><div><font face="courier new, monospace"> 6.21 kb/s total</font></div><div><font face="courier new, monospace"><br>
</font></div><div><font face="courier new, monospace">Connection Times (ms)</font></div><div><font face="courier new, monospace"> min mean[+/-sd] median max</font></div><div><font face="courier new, monospace">Connect: 0 0 0.2 0 1</font></div>
<div><font face="courier new, monospace">Processing: 222 2399 524.8 2581 2641</font></div><div><font face="courier new, monospace">Waiting: 222 2399 524.8 2581 2641</font></div><div><font face="courier new, monospace">Total: 223 2399 524.6 2581 2641</font></div>
<div><font face="courier new, monospace"><br></font></div><div><font face="courier new, monospace">Percentage of the requests served within a certain time (ms)</font></div><div><font face="courier new, monospace"> 50% 2581</font></div>
<div><font face="courier new, monospace"> 66% 2596</font></div><div><font face="courier new, monospace"> 75% 2610</font></div><div><font face="courier new, monospace"> 80% 2616</font></div><div><font face="courier new, monospace"> 90% 2626</font></div>
<div><font face="courier new, monospace"> 95% 2634</font></div><div><font face="courier new, monospace"> 98% 2639</font></div><div><font face="courier new, monospace"> 99% 2641</font></div><div><font face="courier new, monospace"> 100% 2641 (longest request)</font></div>
</div><div><br></div><div style>This is just on my i5 laptop w/ sqlite db on disk, so while keystone is slow, this is even slower.</div><div><br></div></div><div class="gmail_extra"><br clear="all"><div><div><br></div>-Dolph</div>
<br><br><div class="gmail_quote">On Fri, Mar 22, 2013 at 11:47 AM, Chmouel Boudjnah <span dir="ltr"><<a href="mailto:chmouel@chmouel.com" target="_blank">chmouel@chmouel.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Adding Joe email to the discussion since he's not subscribed on openstack-dev@<br>
<div class="HOEnZb"><div class="h5"><br>
On Fri, Mar 22, 2013 at 5:40 PM, heckj <<a href="mailto:heckj@mac.com">heckj@mac.com</a>> wrote:<br>
> We don't have any explicit benchmarking scripts set up and captured for Keystone in keystone's code base.<br>
><br>
> I've been doing benchmarking and live-system analysis, in my case based on post-processing of proxy log output and then reporting latency and processing times for various API calls. My analysis has been mostly whole system and a specific configuration. The logs were generated using various "work the cloud" continual load tests.<br>
><br>
> For single-point direct analysis, you could use any testing system you like that generates HTTP request with specific bodies and headers - grinder, locust, etc. I don't know if anyone working on Tempest has done similar things or not.<br>
><br>
> - joe<br>
><br>
> On Mar 22, 2013, at 7:47 AM, Joe Hakim Rahme <<a href="mailto:joe.hakim.rahme@enovance.com">joe.hakim.rahme@enovance.com</a>> wrote:<br>
>> I wrote a short benchmark script that launches multiple requests to<br>
>> Keystone and measures the response time to get the authentication<br>
>> token. Since I couldn't find anything that does this in Keystone, I<br>
>> thought it'd be a good idea to add a benchmark tool. Swift already has<br>
>> something similar (common/bench.py).<br>
>><br>
>> Since I am new to the community (and OpenStack in general) I thought I'd<br>
>> consult with you here to avoid going in a wrong direction. So here are a<br>
>> few questions:<br>
>><br>
>> - Is there an existing script that does this and that I might have<br>
>> missed?<br>
>> - What features would you suggest I include in this tool?<br>
>> - Should I follow (reuse?) what's already done for the Swift bench?<br>
>><br>
>> Thank you for your answers,<br>
>> J.<br>
>><br>
>> _______________________________________________<br>
>> OpenStack-dev mailing list<br>
>> <a href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a><br>
>> <a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
><br>
><br>
> _______________________________________________<br>
> OpenStack-dev mailing list<br>
> <a href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a><br>
> <a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
<br>
_______________________________________________<br>
OpenStack-dev mailing list<br>
<a href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a><br>
<a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
</div></div></blockquote></div><br></div>