Getting warnings below from our monthly mailshot. Last month I fiddled trying to slow hotmail delivery by adding:
main.cf transport_maps = hash:/etc/postfix/transport fragile_destination_concurrency_failed_cohort_limit = 100 fragile_destination_concurrency_limit = 9 fragile_destination_rate_delay = 1 transport has a list of 28 most popular domains we were delivering to hotmail servers going via the "fragile" transport. transport hotmail.com fragile: hotmail.co.uk fragile: hotmail.de fragile: .... grep fragile master.cf fragile unix - - - - 2 smtp Seems I still sent more email than hotmail wanted (possibly some of that would have been other domains they host not in transport fragile), quicker than they wanted. The server is delivering a couple of emails a second to hotmail servers (which is roughly what I hoped last month change would do), so with only 1100 odd messages still to go, I'm minded that everything is working pretty much as I intended. Is the warning below merely a result of me forcing hotmail delivery to be slow? The error refers to fragile_destination_concurrency_limit being 1, when I set it explicitly to 9. Did I make a mistake? Aside from the odd email shot, the server deals with a relatively small amount of email (after spam filtering at least). So as long as it can handle a handful of messages a minute of other email it'll probably be fine. Aug 6 11:01:56 XXX postfix/qmgr[28205]: warning: mail for hotmail.com is using up 4001 of 6546 active queue entries Aug 6 11:01:56 XXX postfix/qmgr[28205]: warning: this may slow down other mail deliveries Aug 6 11:01:56 XXX postfix/qmgr[28205]: warning: you may need to increase the main.cf fragile_destination_concurrency_limit from 1 Aug 6 11:01:56 XXX postfix/qmgr[28205]: warning: please avoid flushing the whole queue when you have Aug 6 11:01:56 XXX postfix/qmgr[28205]: warning: lots of deferred mail, that is bad for performance Aug 6 11:01:56 XXX postfix/qmgr[28205]: warning: to turn off these warnings specify: qmgr_clog_warn_time = 0
