#22 cache size for already deferred servers

[Neil Burnett]

I am not sure of the status on this, but it was
mentioned a while ago that a problem exists when the
cache of deferred and subsequently processed mail
servers gets too big. Is this still the case?

On a point of logic, is there any point in deferring a
server once it has been proven to retry normally? If not,
this means the cache of servers should be able to get
very big indeed.

[Wayne McDougall]

Logged In: YES
user_id=660239

OK, I'm a bit confused with the point you are trying to make.
There was a problem with resizing the connection cache in
real time. Restarting Fluffy would fix that, but it should be
working in the current version.

Once we have accepted mail from a server, we don't defer
connections again for as long as we remember that site (have
a reference to it in the cache). I try and give priority to
remembering sites that have delivered mail, but this may be
something I need to look at more closely.

But I'm not sure that I'm understanding you, or addressing the
point you are making.

[Neil Burnett]

Logged In: YES
user_id=431787

It sounds like this is no longer a problem. Thanks.

Just for the record, I have configured Fluffy to defer only if a
connection has not been received in the last 10 days. I would
like to make this much longer - preferably forever once the
mail server concerned has actually delivered mail successfully.

I think the problem was that some sites 'fell out' of the cache
within the 10 day period. I will do some checks and continue
this thread if it is still happening. Thanks!