Re: [TSVN] TSVNCache too resource intensive on large workingcopies

Will Dean wrote:
> As Stefan says, we already do some of the really obvious and sensible
> things in the crawler.
>
> I think one of the difficulties is that the crawler is always going to
> generate a lot of disk access, and even if it does that in a
> super-clever fashion which means that it never interrupts anybody,
> people may STILL think 'disk access == computer busy' and wait for it
> to complete.

Can you advise what tools people should use to diagnose what is going on
if they are suffering real problems? It seems that neither you nor
Stefan have any problem, but some other users do. Until we get some
useful data from other systems we can only guess at it and treat the
symptoms and not the cause.

> The crawler is already designed to back-off while the shell is
> actually asking for status info, so that it doesn't slow the shell
> down. It would be good if we could think of other ways of detecting
> that we're competing with other applications, and slow the crawler
> down. Sort of 'thinking aloud', you could imagine doing something
> like watching the disk request queue perf counters and not crawling
> too hard unless the disks were idle.

Won't the negative feedback of its own activity slug it if you do that?

> I don't expect that TSVNCache is using any appreciable amount of CPU
> - it's only disk we're competing for, really.

The big chunk of activity at start-up will go away when the code that
saves the cache to disk and reloads it gets completed.

Simon

-- 
       ___
  oo  // \\      "De Chelonian Mobile"
 (_,\/ \_/ \     TortoiseSVN
   \ \_/_\_/>    The coolest Interface to (Sub)Version Control
   /_/   \_\     http://tortoisesvn.tigris.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tortoisesvn.tigris.org
For additional commands, e-mail: dev-help@tortoisesvn.tigris.org