From: John Doe <jdmls@xxxxxxxxx>
> I have a problem with quick aborts...
> reverse proxy chain:
> 1. Nginx - consistent hashing
> 2. Squid 2.7.STABLE9
> 3. Nginx - fcgi
> 4. php script
> 5. backend (lighttpd)
> For a 7MB test file download (interrupted at around 2MB), I tried:
> quick_abort_pct 1
> then:
> quick_abort_min 16 MB
> quick_abort_max 16 MB
> then:
> quick_abort_min -1
> But squid always aborts the download and does not cache...
> Any idea what could be the problem...?
Nevermind, I might have not waited long enough for squid to finish downloading
the file...
But I still have a question about the "priorities" of the options...
Are they logical ANDs or ORs?
If I use:
quick_abort_pct 1
quick_abort_min 16 MB
quick_abort_max 16 MB
If the dowload is at 2%, but there's still more than 16 MB left... who wins?
quick_abort_pct or quick_abort_max? Is there an order?
Thx,
JD
