chris at simplistix.co.uk
Mon Aug 23 06:50:33 EDT 2004
Shane Hathaway wrote:
>>I have two hypopthesis:
>>1. Some search engine or other spider is ocassionally crawling some
>>non-cached and very CPU intensive pages.
>>2. Something silly has happened that's left each of the clients on app1
>>in single-threaded mode, so more than two simultaneous requests just get
What can I do to rule out the above possibilities?
> Hypthesis 3: it's the network. Perhaps the load balancer or some other box is
> dropping packets. I don't think I saw a delay when I accessed the site
> through an SSH tunnel (to reach a specific app server).
Hmmm, I'm afraid I think I did, but those were connections forwarded
through cache1. In any case, how do we go about proving or disproving
> The front page loads thousands of objects from the catalog,
> so it needs to be
> cached. The rest probably doesn't need to be cached (and mostly isn't.)
Hmm, see hypothesis 1 ;-)
Simplistix - Content Management, Zope & Python Consulting
More information about the Zope-web