I believe the SNFServer also using CPU would hint at there being some spam checking component to the CPU utilization overall. Does the message you are sending have complicated data in it that might require more resources to pattern match? I'd be curious to know if you see the same CPU spikes regardless of message type/content, mailing list addresses, etc. 

the SNFServer is using far less CPU than MailServer, but yes that service is the Message Sniffer for windows which checks the emails. 

The message we are sending via the mailing list system is fairly basic. We send out promo coupons for one of our clients on a bi-weekly basis. They usually contain a header region with few links and a logo, then a body consisting of 2 promo images, and finally a footer with a link to the website, un-subscribe, and various other pieces of information like copyright and address. all the images are linked back to the server, not included in the emails, so the email itself is only about 13kb. but like i said, we do have a list of about 30,000 emails (most of which are doubt opt-in)

Thanks for your follow up on this Richard. My gut tells me we're looking at a pattern matching induced increase in resource usage but grabbing a DotTrace (timeline type) of MailService.exe while the mailing list CPU usage is occurring and getting this over to our support team might allow us to identify what specifically is generating the increases. 

I am running into the same issue and wanted to see if anyone else has been experiencing high CPU. Mail service for me is currently running 95% of my CPU and this was not like this before upgrading to Build 9182 (Feb 20, 2025).

I was getting ready to submit a ticket.

We do mailing list mailouts every 2 weeks, so i won't be able to run the dotTrace until then to produce a log. We are also with a provider, so we don't have login's to submit an office ticket (i have to collect everything and get them to submit) but if you can get the information before us, please do submit a ticket, it would be appreciated! hopefully we can get to the bottom of this.

If your spool thread count is higher than your hardware's actual cpu/thread count, that would also explain the CPU usage. I'd recommend adjusting the spool thread counts to reduce pressure on the CPU, if this is the case.

Would that be this setting: https://portal.smartertools.com/kb/a2708/increase-the-total-number-of-emails-allowed-in-the-spool.aspx?KBSearchID=938533

If i'm not mistaken, we have the default settings for spool threads, we have a fairly standard intel xeon gold 6326 on our server as it's not HIGH load like some others.

OK, we finally did another mail out from Mailing List, and right now the CPU is spiked at almost 100% for over and hour as it processes... 

We installed the dotTrace, but not sure exactly what we are looking for... i was given this screenshot if it helps.

*edit 

i checked our admin, and currently we have 9,405 items in our spool and 5,813 waiting to be delivered, out of just over 27k that were apart of the mailing lists (we sent out 5 different lists with various amounts at staggered times) when i check that domain our "outbound messages per hour" is set to 10,000, the contents of the email only come out to about 80kb, we use linked images instead of embedded images.

I believe what they'll be looking for Richard is a timeline type trace of MailService.exe for at least 3-5 minutes or so. That should give us a few into what is happening under the hood.

i thought that was what i did in the screen shot lol the Timeline i have just shows a steady stream of ".NET TP Wait" with values at 7,341 MS, my guess would be hundreds in the short snapshot i just took again. and when i look for details of the call tree, i get https://miramar.ca/UploadedFiles/images/test/dotTrace2.png

Richard, I started a ticket off this thread. If you can send us the actual dotTrace, that would be great.