2

I would like to create a list of default times that mail server will try to deliver email to other email server before it sends information Unable to deliver back to sender (I mean temporary errors here like network down, server down, no connection, non permanent error messages from server).

I'm interested in broad spectrum of email solutions (possible clients) that may send us emails to write documentation before introducing Exchange 2010 in our network as a mitigation plan so that we could safely assume 8 hours downtime (in case of problems) won't cause any problems in terms of email delivery that management may .

An example I've found for GMail (not confirmed)

Gmail treats mail to bad domains as a temporary failure (because it may be caused by a DNS issue). It will warn you after 24 hours that it hasn't been able to deliver the message, then keep trying for two (if I remember correctly) days before sending you a final failure message.

I would be interested in default settings of multiple servers and widely available free accounts services. I do understand that some administrators may have tweaked the settings but this information would give me nice overview.

Just to add as a side note. I will have 2nd MX server and redundant solution and this information is to give an overview of crisis situation where everything goes down and we have to fix it. Don't want management to panic during time we fix it.

MadBoy
  • 3,725
  • 15
  • 63
  • 94

1 Answers1

4

"As long as the local admin configures it to"

A warning after 24 hours is pretty much a de-facto standard (vanilla Sendmail is 4 hours if I remember right).
Retrying for 4 days is pretty typical (it's what plain vanilla Sendmail does).
With most MTAs the local admin can override both of these settings and make them whatever he or she would like. Freemail providers can change the settings at any time and for any reason.

You can generally assume that outages of up to 8 hours will not cause problems for MOST mail services (the mail will be queued at the sending server), however the correct thing to do is to set up a secondary MX (a machine that accepts and holds mail for your domain for longer than 24 hours in the event that your main server is offline). Google for "Secondary MX" or "Backup MX" for a list of companies. I can vouch for the one I use - Twisted4Life - as being quite reliable.

voretaq7
  • 79,879
  • 17
  • 130
  • 214
  • I do know about secondary MX and we will plan redundant solution (Exchange 2010 with DAG) with DUAL LINK and possibly 2nd location but I'm trying to state that there will be huge consequences if there will be disaster lasting some hours. – MadBoy Aug 23 '11 at 15:07
  • @MadBoy - If your secondary MX is not geographically and topologically separate from your main environment your are doing it wrong. e.g. I am in New York, so my secondaries are in California and the UK. Chance of a disaster affecting all 3 regions: Negligible. – voretaq7 Aug 23 '11 at 15:10
  • Our second location is about 50km away from us. – MadBoy Aug 23 '11 at 15:13