Re: [RFC] Breaking forwarding loops

On 01/10/2011 03:02 PM, Henrik Nordström wrote:
> fre 2011-01-07 klockan 11:56 -0700 skrev Alex Rousskov:
>
>> We hit another flavor of the same bug with Squid 3.2: either the complex
>> network setup allows loops for forwarding ports or the flags.transparent
>> test is failing because we do not set that request flag for transparent
>> ports if certain transparent redirection actions fail.
>
> If someone want to clean this up then they have a +1 from me on making
> redirection lookup failures terminal on transparent/intercept ports.

I seem to recall reports that "the proxy was still working fine despite
the lookup errors" so we should be careful not to disable useful
functionality. There is probably more than one way to lookup the
destination, and as long as one way is working, we should probably keep
going.

>> Does anybody know why the original code did not break all forwarding
>> loops, including loops for regular forwarding proxies?
>
> it does. It only is a bit harder when interception is used. When
> interception is not used it's assumed a loop can be broken by going
> direct making it a soft "invisible" error only logged in cache.log

Do you think we should break all loops because it is difficult to be
sure which ones are going to be "fixed" by going direct? The alternative
is to try harder when identifying the situations where a loop can and
should be broken.

Please keep in mind that unbroken unfixed loops create a lot of problems
because they generate huge logs fast and produce huge requests that
stress fragile Squid parsing code.

Thank you,

Alex.
Received on Mon Jan 10 2011 - 23:19:30 MST