Forums » SEPHP Help and Tips

Notifications not working - no member getting any emails

    • 14 posts
    October 14, 2017 12:24 AM EDT

    hi 

    despite members having everything ticked - they are not getting any email notifications from the site. I manage the domain on my server and the emails aren't even sending, it's not that they are getting blocked they don't even show as going out? 

    It looks like the last time it ran was June? 

     All default notifications are set to 'on' 

    any suggestions would be appreciated.  

    it didn't make any difference when I changed to default mail or smtp mail - neither work 


    This post was edited by Maria at October 14, 2017 1:38 AM EDT
    • Moderator
    • 4694 posts
    October 14, 2017 5:36 AM EDT

    The images are too small to see. Even opened in a new tab. 

    Try different methods of the task scheduler settings. Perhaps the server is blocking it from running. Check error logs with your server and in the log browser in admin.

    I use Sendgrid for my site transactional mail. On that note, it reminds me about the mail ports. I think error logs should show issues with that though but I'm not 100% sure. Hosting is my BryZar partner's area of expertise.

    • 14 posts
    October 15, 2017 2:27 AM EDT

    thanks Donna, I re-generated the trigger access key in the task scheduler settings and that seems to have sorted it out.

     

    thanks for pointing me in that direction

     

    cheers

    Maria

     


    This post was edited by Maria at October 15, 2017 2:27 AM EDT
    • Moderator
    • 4694 posts
    October 15, 2017 4:53 AM EDT

    Great! Glad it's resolved for you.

    • 6 posts
    May 11, 2018 10:53 AM EDT

    Im having a similar issue, which no one can seem to resolve.  Ive been told from my hosting company, TMD hosting that there is an issue in the social engine core email utility.  The site sends welcome emails ok, but when you are logged in and click the invite link and put in emails to invite to the site, the emails are not being sent.  TMD says all the logs are even empty, which means the utility is not even sending them to the server.  Im on a really tight budget with this so im not able to get the SE support yet, but it seems to be a core issue, so i was wondering if thats still my responsibility to have to pay for?  Any help on this is greatly appreciated.

     

    Jerome

    • Moderator
    • 4694 posts
    May 11, 2018 11:23 AM EDT

    I've not had issues getting emails to send out from the site as I tested it this morning for a different bug report. If you need us to check, yes you need to purchase support. Best thing to do is check the knowledgebase for various tutorials for email and try that first if you want to test everything.

    • 31 posts
    May 11, 2018 12:58 PM EDT

    Glad to see you were able to resolve. We had a similar issue but it was due to outbound SMTP traffic being blocked on the port we use with sendgrid. We opened the port and all was good.

  • May 13, 2018 12:01 PM EDT

    I've had issues with this in the past. I've tried several solutions, including editing the database values of the actual number of seconds, to edit the frequency of times tasks run in the task scheduler via curl. As well as fixing up max execution and max input times to mirror the setting -1.

    What I've discovered, is that there are times, in Social Engine, where the task just stops running. It times out on its own, even when it's set never to do that. Curl for some reason, will randomly disconnect the task of running mail. by default, Social Engine makes the mail task run every 15 seconds. I changed that to run every second. That way the mail is sent out as instantly as possible. I also updated some of the other tasks, including decreasing the number of times jobs are checked, from 5 seconds, to 1 second as well. I made updates to how long a task can run without fail, to ensure that on the hour, every hour, the task is run again through enforcing max input time and max execution time was set to 3600 seconds in php extensions, php.ini configuration file as well as the MYSQL database where the values to update social engine files to edit the frequency values for the task scheduler itself are stored. You can tell, this is advanced tweaking here and if you don't know what you're doing, I don't blame you.

    The bottom line is this. Preventing MYSQL task scheduler from timing out, refusing to run a task after so long of running it successfully. I got it to work, but the time frames will be different based on your community size. If you have a community of less than 100 members, and don't send out mail that often, you can get away with more time. But if you have a large community, with lots of updates, you'll want to ensure that notifications don't delay in the back end. Including the sending out of newsletters, invitation emails, and other notifications tied up in task scheduler settings managed via either CURL or Cron. Both of whom, time out randomly. So take notice of this, and correct the timed out status of this by looking up the best time to enforce a curl or cron refresh. You'll have to do this manually by ensuring you temporarily edit the values of the timed out processes in question, change itself automatically to unstuck itself from having to wait that length of time to try again. This is why the closer to real time you can get the values in the database, the better off you'll be. When the values in the database match a reasonable length of time in php settings a script can run in the back end, you should be all set. In theory, that's how I was able to solve this problem on my end of no emails being sent by social engine. By altering the value of the effected processes which timed out, to a minute, and then returning it to its original setting using the database values in engine4 > tasks I believe.