Re: Connection timeouts and getaddrinfo

[Dave Brolley <brolley@xxxxxxxxxx>]

On 05/03/2016 07:26 PM, Nathan Scott wrote:
> Hi Dave,
>
> ----- Original Message -----
> > On 05/03/2016 02:26 AM, Nathan Scott wrote:
> > > [...]
> > > Not sure what the correct behaviour should be here - thoughts?
> > > Seems like its probably not doing what users would expect atm.
> > The delay is applied during the call __pmSelectWrite(). One thing we
> > could try would be to open a socket for each address, use the select to
> > wait on all of them at once, and choose the one that's selected. If the
> > timeout expires, then we can assume that they all timed out and we will
> > have applied the timeout once for all of the addresses. I can't think of
> > another way to apply one timeout while trying all of the addresses.
> *nod*
>
> > The downside is that PMCD will see several connections, some (most?) of
> > which will succeed and then be abandoned.
> Yeah.  pmcd also has an optional connection-count-limit feature that we
> might bump into this way ... but I think those downsides are probably
> outweighed by having the timeout applied in a way people would expect.
One other possible pseudo-solution: If the first address times out, try 
the remaining addresses with a much smaller timeout, since they are 
likely, but guaranteed to also timeout. This would keep the total 
timeout time close to what was requested while avoiding the simultaneous 
connection attempts.
Dave