Re: Fetching mail repeatedly but without duplication.
Matthias Andree <matthias.andree <at> gmx.de>
2008-05-19 16:39:03 GMT
Pratap Subrahmanyam schrieb:
> I'm having a small issue that I need to workaround.
> All my mail arrives and resides in an MS Exchange 2007 mail box. I then use
> fetchmail to get it from there, via IMAP4, send it through procmail to have
> it filtered into folders etc, and then read it via mutt. It all works mostly
> just fine and I like my setup a lot.
> Recently, due to some IT screwup, all the mail that I had incrementally
> fetched over the last few years have been lost. Thankfully, I left all my
> mail on the server!
> Now, I have about 61K emails on the server
> I decided to do a "fetchmail -ak". But, it has failed the last 7 attempts,
> some time after the low teen thousandth email, once after 13K once after 17K
> emails etc. And it takes a long time to do this. Each time it fails, I "rm
> -rf ~/Mail/inbox" and try again. But, it seems to me there has to be a
> better way. Each time the error is some socket failure. I totally believe
> this is probably a flaky mail server problem and has nothing to do with
> fetchmail itself. Still, I need to find a way around it.
> Here are some specific questions:
> a) is there a way I can tell fetchmail to fetch emails numbered 1 - 1000,
> then 1001 - 2000 etc? The chances that there will be a socket error while
> bringing in a 1000 emails is, let's see, about 60 times smaller than if I
> asked it to fetch 60000 emails. I like those kind of odds.