Re^11: Starman problem

In my experience Starman is solid and highly reliable, I suggest you add some profiling mechanisms and logging to your own code and try to isolate what is going on and what it is doing when it causes Starman the problems you describe.

Comment on Re^11: Starman problem

Replies are listed 'Best First'.
Re^12: Starman problem by Logicus (Initiate) on Feb 15, 2012 at 09:27 UTC
Thanks, that's what I needed to know (and suspected). The problem hasn't recurred in the last 2 days, and only seems to crop up with a frequency of about once or twice per week, where CPU usage spikes to about 18-20% for a short period before apparently fixing itself back to it's usual level of well below 2% I've put in logging mechanism now so that if a page takes longer than 0.02 seconds to render it records which domain and what action of the system was being called, hopefully when the next spike occurs that will be enough information to single out the cause. Once again thanks for the second opinion, it was useful!	[reply]

Replies are listed 'Best First'.

Re^12: Starman problem
by Logicus (Initiate) on Feb 15, 2012 at 09:27 UTC

Thanks, that's what I needed to know (and suspected).

The problem hasn't recurred in the last 2 days, and only seems to crop up with a frequency of about once or twice per week, where CPU usage spikes to about 18-20% for a short period before apparently fixing itself back to it's usual level of well below 2%

I've put in logging mechanism now so that if a page takes longer than 0.02 seconds to render it records which domain and what action of the system was being called, hopefully when the next spike occurs that will be enough information to single out the cause.

Once again thanks for the second opinion, it was useful!

[reply]