Being beta, we'd like to roll back to the previous beta build, is this a safe action though?

I was wondering if with the latest public "production ready" beta, it was now ready for production so we can update our main server, but it looks like it's not the case :(

[rant on]

There are bugs in our current production running pre-MAPI (7242) that our customers are waiting to be fixed. Some are fixed in the beta but we can't switch to it as it seems in brings more problem than solutions.

We pay for maintenance and support, but fixed bugs aren't pushed back to the "stable" version and beta seems to have still it's amount of issues.

[rant off]

I hope you will be able to rollback to the previous beta.

Yeah... no... stay far away from the MAPI beta right now.

POP3 is broken and has been broken for over a month for I'd say a good portion of our customers, though this latest build was supposed to have that fix, but instead of POP3 being fixed, the entire mail server is crashing now. Though, the POP3 debug ID 20191227 is no longer showing exceptions, so maybe it is fixed :)

IMAP has issues on and off for over a month for quite a few clients, ranging from UID changed errors to odd header errors

MAPI we cannot use at all because it crashes our IIS worker process

Oh okay, thank you for the feedback.

The POP3 issue is maybe fixed, or not but the server can't run long enough to log the error...

Let's hope it's the first answer :)

For MAPI we don't consider using it yes and propose it to our customers.

We want first to have the other existing protocols/services running smooth. Even that we are paying for it because we use EWS for some customers (mac mail on osx and emClient).

I did ping Matt and Robert, hopefully we can get a fix for it shortly or at least if it is safe to rollback to the last beta build or not.

I hope you'll have an answer soon, but it's the weekend, I would bet more on the rollback.

If i'm not wrong the rollback problem is between pre-MAPI and MAPI as files content have changed and wouldn't be recognised by pre-MAPI builds.

Yep, just making sure something has changed between betas that could cause issues. To be fair though, ST has been really good at responding to me when we have issues. So, here is to hoping we can get it diagnosed and fixed up.

I'm on and will be evaluating for issues. I'm checking our logs for any anomalies and I asked Chris for some event logs those should have the crashes in them.

EDIT: It appears Chris is somewhat stable right now and hasn't crashed in a couple hours. We'll be combing through logs looking for anything bad. If we notice anything or if we get any other reports of a crash we can evaluate further steps. I also gave our server a check and noticed no issues since our internal update we did yesterday.

Yes, been almost 3 hours since the last crash, though, I dont have the warm fuzzies and I am scared to walk away from my desk, I know just as I get home and crack open a cold one, SM will say "Oh no, hold my beer" and take a poo.

Matt, I sent you some exceptions from the 111 log, we are getting close to 100 exceptions in that file in the last 10 or so minutes.

Looks like there really is an issue with the new beta build - Chris is not the only one who has to report service stopping/crashing!
Right now I noticed that our SmarterMail-server stopped/is not responding anymore.

@Matt: here the public beta build 7334 was running 5-6 hours, but then the Windows Service stopped - did not happen with the beta builds before this one!


What kind of logs do you need - or what Debug Log IDs should I add for logging?
I will reboot server and see how long SmarterMail will stay running this time.

In case it continues with SmarterMail-Service stopping/crashing I will need to downgrade to previous public beta 7327 before start of the week.

Sorry Thomas, glad we arent the only ones though :). Sometimes I feel like we are alone with all of our issues lol.

We had to roll back last night, it crashed after about ~8 hours or so of running, then after the restart it crashed minutes later, we rolled back to the previous beta build.

We're going to have to roll back too.  Also getting service crashing every 20-30 minutes and today is an extremely high volume email day in our world so it's going to be a long day.

Now our SmarterMail/server+service running build 7334 has uptime >30 hours without crashing but this was during weekend with almost low traffic/low usuage.

I will keep an eye on it what happens during Monday/start of the week, when our mailserver will be more active/being used/accessed by our employees.

Monday morning 05:54 the SmarterMail-MailService.exe stopped/crashed again :-(
Uptime around 39 hours - but this was over the weekend so less Mailserver usuage than usual.

Looks like there is a stability issue with actual beta build 7334 - hopefully we will get over today without downgrade to 7327, perhaps SmarterTools will have an updated custom build available at end of this day???

In case you need specific logs please let me know what could help debugging/solving these fatal issue.

I sent some of the memory dumps we had to them, hopefully those help figure out what the issue is. Perhaps if you have your MailService.dmp files you can provide them as well, maybe one of our dumps will shed some light on the issue.

09:02 the next crash - this time only three hours uptime :-(
will try to activate memory dumps

Chris those dumps were immensely helpful and immediately pointed to the issue. We are patching our servers right now and we'll likely be doing an update here tonight. Thanks again for reporting these quickly.  

@Matt:
what kind of logs/dumps do you need,

what steps must I do to create helpful dumps? using a special tool?

and in case you have a fixed build available please send me the donwload-link so that I can give it a try in our environment. Thanks.

We just pushed another Beta release with this issue fixed.

Note to self - don't install a Beta that's released on a Friday again :)

Agreed, I wish they did releases on Monday instead of Friday, that way the issue can be fixed up that day or the following day.

until now build 7338 has fixed the "service-crashing"-issues - I did not notice any service-stopping today.

and yes, please no new release/no new build on Fridays was very often suggested,
due to the new 4days-work-week at SmarterTools better updates not on Thursdays and not on Fridays!
So it should be easier for the devs to react and supply a fix :-)

So far, no further crashes and with today's update, hopefully we are stable until we can diagnose the MAPI issues.

Thanks guys!