Fadi Mansour <[EMAIL PROTECTED]> wrote:
> While browsing some pages from the wwwoffle cache, I keep getting the page
> that is usually presented if another process is writing to the same page
> (this is in case two wwwoffle instances use the same cache dir). But in this
> case the other wwwoffle instance is surely not writing to this page.
>
> I tried then to solve the problem by deleting the offending page. The
> problem now is that whenever I try to access that URL, I get a message that
> it's not in the cache (I have already deleted it). But when it's time to
> actually fetch the page, I found out that it wouldn't get fetched, looking
> at the log file I found that wwwoffle reported that it's "already fetching"
> that particular page, which is not true.
>
> Any idea what's happenning. Andrew said that using two instances of wwwoffle
> on the same cache was not very common (so to say), but the question is, how
> does wwwoffle know that it's "already fetching" a URL (or that it's being
> written by another instance).
Even though there is only one WWWOFFLE program running it will start
several processes to fetch the pages for you. It is possible for two
of these to be fetching the same page at the same time.
To stop this causing a problem and reduce the bandwidth by stopping
the same page being fetched twice I added lockfiles. These are files
that are created in the cache that indicate that a page is being
fetched.
A lot of people did not like this and in recent versions of WWWOFFLE I
have made them optional. By default they are turned off. To make
sure that this is the case set the lock-files option in the Options
section of the configuration file to false.
--
Andrew.
----------------------------------------------------------------------
Andrew M. Bishop [EMAIL PROTECTED]
http://www.gedanken.demon.co.uk/
WWWOFFLE users page:
http://www.gedanken.demon.co.uk/wwwoffle/version-2.6/user.html
