These are chat archives for ManageIQ/manageiq/performance

2nd
Dec 2015
Alex Krzos
@akrzos
Dec 02 2015 16:20
Screen Shot 2015-12-02 at 11.19.23.png
@kbrock 5.4.4.2 on the top, 5.5.0.12 on the bottom
x-large satellite 6.1 configuration management refresh
much more cpu usage on 5.5.0.12
Alex Krzos
@akrzos
Dec 02 2015 16:25
634s (5.4) vs 890s (5.5)
for refresh
(sorry for the issue :smirk: )
Dennis Metzger
@dmetzger57
Dec 02 2015 16:28
@akrzos thanks for shining the light ...
Alex Krzos
@akrzos
Dec 02 2015 16:29
Probably not a priority right now but figured I'd mention it, since I was able to add that provider (Sat 6.1) into my benchmarks
Keenan Brock
@kbrock
Dec 02 2015 16:30
@akrzos are you trying to depress me?
Alex Krzos
@akrzos
Dec 02 2015 16:31
haha
Dennis Metzger
@dmetzger57
Dec 02 2015 16:31
no, he’s trying to distract you
Alex Krzos
@akrzos
Dec 02 2015 16:31
I thought I wasn't trying hard enough :smile:
Keenan Brock
@kbrock
Dec 02 2015 16:31
no - you're distraction is working :)
Alex Krzos
@akrzos
Dec 02 2015 16:31
my bad
Keenan Brock
@kbrock
Dec 02 2015 16:32
lol
nope
You're keeping me motivated and excited
my wife is not too happy with you though. (I'm working around the clock on all the neat performance fixes we can make)
akrzos @akrzos goes back to his performance cave
Keenan Brock
@kbrock
Dec 02 2015 16:33
ha!
was speaking with yaniv
I mentioned all the cool influxdb stuff you are doing
Dennis Metzger
@dmetzger57
Dec 02 2015 16:35
I liked how you snuck @akrzos into the conversation
Keenan Brock
@kbrock
Dec 02 2015 16:35
heh
Alex Krzos
@akrzos
Dec 02 2015 16:35
Still got it running, I think I need to update it (influxdb), but last time I tried, it lost my db, good thing I snapshoted the disk
Keenan Brock
@kbrock
Dec 02 2015 16:35
I'm living vicariously through @akrzos - he gets to use influxdb
also living in shame as I have yet to export events to influxdb
Alex Krzos
@akrzos
Dec 02 2015 16:37
I'm excited for getting a 5.5 release with the proctitles
I deployed a master appliance to see them
was going to add collectd to verify we have what we need for collectd to separate the workers
Keenan Brock
@kbrock
Dec 02 2015 16:38
@dmetzger57 found the page for graphite tools: http://graphite.readthedocs.org/en/latest/tools.html
Greg Blomquist
@blomquisg
Dec 02 2015 20:29
allow me to derail again ... as we start looking to really support containers, and think about the scale of containers, I worry about our ability to scale with those numbers ... @akrzos can you start looking at getting out in front of that train with some of the openshift guys?
maybe start looking at the numbers they're thinking about for their environments ... that's probably the top end of the scale we'd be concerned with
Alex Krzos
@akrzos
Dec 02 2015 20:32
@blomquisg Yes, it's definitely been on my to-do list
i have an OSE environment deployed, just need to get some tainers running on it but should also sync with those guys on the expected scale
Greg Blomquist
@blomquisg
Dec 02 2015 20:33
'tainers
haha
Jason Frey
@Fryguy
Dec 02 2015 20:33
makes me think of tainer-tots
Greg Blomquist
@blomquisg
Dec 02 2015 20:35
how can we work "pods" into this?
Alex Krzos
@akrzos
Dec 02 2015 20:39
OSE has pods, when i first spawned the OSE instance I have it wants me to make some pods
Jason Frey
@Fryguy
Dec 02 2015 20:40
tainer-pods?
Alex Krzos
@akrzos
Dec 02 2015 20:40
just really need the hour or two to spend on it to create a simulated running ose environment
Greg Blomquist
@blomquisg
Dec 02 2015 20:40
ooh, do they have something where you can simulate the pods/containers?
like vcsim?
Alex Krzos
@akrzos
Dec 02 2015 20:41
Was hoping to do that with a single vm, may not be a full on simulator but pods with containers running nothing but maybe sleep statements or something