Page MenuHomePhabricator

performance of vagrant in labs instances
Open, MediumPublic

Description

I was wodering about slow performance on my labs instance.
http://drmf-beta.wmflabs.org/wiki/DRMF
There are a lot of 503 errors for example if I try to load
http://wikidata-drmf-beta.wmflabs.org/wiki/Q24

htop does not show a high usage either.

2016-07-22.png (1×960 px, 131 KB)

Do other labs users have similar problems?
Is there a best practice guide on digging for the performance bottle-neck?

Event Timeline

@Physikerwelt I am experiencing a similar issue on my instance http://whgi.wmflabs.org/ .

What I experience is that intermittently I will see my python and java jobs slow to a crawl (they don't log) at about 1/100 speed. However Top won't show any abnormal CPU utilization. Same issue \. It's as if the clock speed on the CPUs was just suddely turned down. In these cases even bash is slow to return results.