Sending process halts on not responding mail server

Hi, Christopher
Question: What should happen, if the sending process stumbles on a problem with an email address? Because this is what happened already twice to me. The process was halted because one mail address was seemingly not reachable or not replying, and sending would not continue until a Clear Sending Status was envoked manually.

There is a setting in Options -> Optional Features Stop Newsletter Sending on Malformed e-mails and this is set to set as unconfirmed and continue sending  First I thought this would remedy he situation, but no.

It doesn’t help at all. The problem is neither a bounce, nor a malformed address, but rather a server that cannot be reached, perhaps only momentarily for whatever reason. I feel this is a bummer as I never encountered such behaviour when I was still using Acymailing. It’s alright to get an admin mail when something like this happens, it would be alright to mark the subscriber as unconfirmed, but after that then NS Pro should continue sending rather than sitting idle until somebody comes along to rectify things.
So, I'm open to suggestions.

Cheers, Chris

Hi Chris,

First of all, you can just click "Send" from your newsletter manager, at any point, and the newsletter will just continue sending from where it stopped. (unless of course if it is already sending at the moment)
There is never a need to clear the sending status, unless you want to start sending over.

If the sending has halted mid-sending and the newsletter can't be sent, then your server reached a limit and probably killed the php script. This shouldn't happen, so if it happens you should probably change your sending settings.
When it does happen, the newsletter seems as still sending, but it is not actually sending. NS Pro will assume the script has stopped working after 2 hours, and automatically mark the sending as finished and continue sending.
(All this assuming you have scheduled cron job sending)

The manual way to fix a halted sending (earlier than the automatic two hours) is by going to the sending log, and entering the last sending and clicking "Mark as Finished". Then the newsletter can be sent again.
These procedures may seem uncomfortable at first, but they are all security measures against bypassing a subscriber or sending newsletters twice.

If a mail server is unreachable, the sending process will simply stop and try again when triggered from the cron system. If the mail server is still unreachable, then this is not a temporary problem, but one that needs the administrator's attention. The subscriber should probably be unconfirmed or moved to a separate list.

I hope this all makes sense.
Thanks again!

I see, I have to explain in more detail what happened:

• I am sending my newsletters using a cron on my own hosting (a stripped-down version of the example cron you have written)
• I actual sending is done using not the SMTP service of my provider, but Amazon SES. SES claims a limit of 13 mails/s
• So I have configured NS Pro to send mails in batches of 40 mails
• The cron job runs every 5 mins
• There are around 400 subscribesr in that mailing list, so the job should finish in less than 1 hour
• But still within the first batch of mails, the process was halted and stayed stuck

To test the setup I had created a bunch of around 30 artificial mail addresses (simple forwarders called test-1@, test 2@ and so on and sent sample mails out, without encountering any problems.
As explained above I then sent out my first real mail, it got stuck, and it stuck within the first batch of mails, and it happened after 38 mails werde sent (in the second newsletter after 30). The process halted, the cron was still running and now I started getting mails from the system, reporting that the previous sending session was still active and sending could therefore not continue. I am not aware how to recover from this situation, other than clearing the sending status, finding out to which subscriber the last mail went out, and then setting the address coming next to unconfirmed, before continuing. That resolved the situation. To complicate matters further, when all was sent (and there were no further hiccups, every batch of 40 mails was sent), I tried to send the mail manually from the tool-box to the subscriber in question. Now, that went out without complains.

I don't know what was causing this problem, and it happeed again sending the second newsletter. That's why I am asking. Perhaps you have a suggestion. Newsletter Max Time Limit is set to 240 - I don't think that's the problem. Apart from that I cannot see any settings that would cause a problem

Hi Christopher,
I think the problem indeed lies not with NS Pro, but with SES, there seems that a limit sometimes strikes and causes the issue. I reduced the amount of mails sent in a batch and decreased the interval betweed cron triggers, and everything was sent fine. So, no further action on your part is needed. Problem resolved.
Thx for your thoughts
 

That's excellent news Chris!
Can you please share the final settings that worked for you?

For example: X emails per Y minutes?

Christopher,
I'm still in the trial-and-error-phase. I originally set NS pro to send 40 mails every 5mins. That worked basically, but there were hiccups, where SES got stuck and would not continue. SES says that there is a limit of 14 mails/s, but during testings that was never a problem. But in real life things seem to be different. 
Anyway, I tried the last newsletter with a setting of 28 mails and the cron engaging every minute. That worked without problems. The next time, I will increase the amount of mails to see what will happen.