Less common advice for mid-session CTD's

Discussion Hub General Discussion
, , , ,
1

So I noticed we are still talking about CTD’s in the forums, and I wanted to share some less common suggestions without it getting buried in a thread. So please pardon the quasi bloat post. :pray:

Cache

Have any of you tried precaching your flight region? Most of my rare CTD’s come by way of what I suspect is a hitch in downloading map content - particularly photogram. I have 1gb bandwidth, NVME SSD, etc - but I have observed hitching as content loads into the sim that has escalated into a CTD. It was at least noticeable enough for my programmer brain to kick in. So rarely recommended advice #1:

1.) Try precaching your flight route. Full Quality.

The follow on to this could be an issue with the hard-drive map cache filling up and beginning to overwrite. I have no idea how to test this easily, but I can imagine it could cause a hitch and CTD if there is a memory management bottleneck during scenery loading. (I have seen the sim do this, but it was in 2020 and during downloading marketplace content updates. this could be out of date, but if your stuck … try it!) So, #2:

2) Keep plenty of clear cache, maybe dump the cache periodically and see if that helps.

Further musing … you may not have enough swap space (or RAM) on your drive outside of the cache itself. Your computer has to put all of the virtual planet somewhere, so give that a check.

Autopilot

Are you using the default Autopilot? Example: I managed to find a delightful CTD when on AP and tracking a secondary VOR overlaid in my HSI. I discovered first hand that the default AP has some sneaky bugs. So :arrow_right: Try the NXI beta that is free on the Marketplace. It is a much improved experience and I have yet to experience a CTD flying on AP after installing the beta. If you know your way around a Garmin (maybe a bit too well) and you may have encountered a mysterious bug at the same point in your favorite flight repeatedly, ask yourself if you’re on default AP. That’s what happened to me, and nudged me into trying the NXI beta. No more crashes, No regrets. So rare suggesting #3:

3) Don’t use the default AP if your fancy. Use the free NXI beta on the marketplace.

AMD?

I have had CTD’s extremely rarely since launch, on two separate pc builds. I’ve pondered this one for the entire 2 years, and the only thing I can notice is that I am an oddball with a pure AMD rig. Twice. My underspec machine rarely crashed on an ancient ATI card. The current rig is on a 6900xt. I have to wonder if the fact that my hardware more closely matches the latest XBOX is an indicator. Not really sure here, but if you happen to have an amd card lying around and feel like hardware experiments - give it a go.

Anyway, try these out if your stuck & frustrated, :sunglasses: and yes … you have a clear community folder. :laughing: If none of these work, you might still take a look at what your common mission is. You may have found a fresh bug.

And what do we do with fresh bugs? Report them, so someone else can fix them. :rofl:

Fly safe! :wave:

[Edit] Nice call out in the thread. Check your libs!

4) Run Windows Update for the latest .NET libraries

2 Likes
2

MS Asobo should warn os this before simulator updates:

3

Why should they?

Asobo is in no way related to the Windows or .Net team so they don’t know when an updated will enter the public ring.

Even if all is under some Microsoft lable this doesn’t mean all teams know what other teams are planning or doing.

4

To be honest they really should have a good idea or they are in the wrong job.

5

I am in such a job actually, so I can tell that a third company like Asobo won’t even be able to view a kanban board on Jira when it comes to Microsofts internal teams other than these they are working with directly.

Asobo has a management SPOC (Neumann) within MS and he basically also could only see stuff related to the gaming sector.

If you have an actual idea on how to come around political (privacy as we are talking about EU vs. US) and company secret topics (Asobo still is a different company than MS so Asobo only has restricted access to projects within MS) without ending up violating any NDAs feel free to provide that :wink:

6

I understand that but really they should have basic info as part of their contract or is this not basic?

1 Like
7

Well this took an interesting turn. :rofl:

With respect, the teams that build dependency libraries publish their roadmaps. Breaking changes are usually announced with lead time to allow for patches. There is usually nothing secret about a library team’s roadmap. So Asobo would have an ability to know a change is coming if they read the roadmap notes.

Huge Caveat Inbound …

More likely, the dependency isn’t one of MSFS’s direct dependencies, but something maybe the XBOX client is using. So the dependency is an indirect, extended dependency. Now that is hard to know about. The argument about access to codebases become relevant for these nested dependency issues. Asobo can’t know if XBOX client team or MSStore team are using x class to access y service. Things can break a little unexpectedly in this case.

So yea, not exactly cut and dry that. :wink:

The amount of distributed teamwork that makes this sim work is most impressive despite the imperfections … at least to me anyway. :laughing:

8

I am totally amazed at the Asobo team and the work they do and no matter how much they get paid. I have worked with computer system rollouts and for some very big projects. The amount this team deal with in terms of ongoing changes, fixes and improvements is huge. Anyone who criticises them should spend some time doing the job. It is a moving target and some who say we have not got what we payed for do not realise what it really costs. We may not have every single thing promised but “WE” have demanded they do other stuff “WE” think is more important. This is somewhat off topic sorry.

1 Like
9

Agreed. Their stakeholder meetings Developer QnA’s are fantastic. :grinning_face_with_smiling_eyes:

10

As expressed in a thread I opened up yesterday, beware that CTD’s can also be server-related! Not many of us would have even considered a ‘link’ between the two things, prior to the SU9 update.

So all these measures mentioned here client-side may be redundant until any ongoing server issues are fixed. Just making you all aware of this so you don’t spend hours trying to remedy a CTD at the worst possible time (like I did).

I’ve not had a single issue from yesterday onwards, despite my efforts to “recreate”/diagnose a CTD.

11

Indeed, this can happen and was the genesis of suggestion #1 to precache the world. :face_with_monocle:

The theory goes something like this:

What data types & interactions is the server sending me?

  • World Data
  • Updates on other aircraft
  • Weather
  • and so on…

The reliability of MP permanence is shaky, but nothing crashes when my buddy phantoms in and out of existence as we fly. So I am less likely to suspect the network packet layer itself.

Now the Gb’s of world data? Could totally cause a crash. If AZURE has a hiccup, or your ISP drops your bandwidth for a few seconds, etc the sim might enter an unrecoverable state due to packet delay/loss.

Weather? Maybe, for similar reasons, but you can add the enormous data task of integrating in the various weather data streams into a single coherent model we receive. So any hitch or error in the weather data model formation could trigger a CTD bug.

Obviously, I don’t have the source code and I am speculating. In stands to reason though that as a troubleshooting step: turning off live weather and precaching all of your flight region before entering a flight is a solid way to reduce server load. It might be that your geographic location and ISP are part of your CTD problem. Caching as much as possible and turning off live weather will reduce the effect of inconsistent/unreliable network performance on your sim.

Fly Safe!

[edit]
You know … the whole weather thing really has me thinking. If MS/Asobo are ingesting raw METAR data and other public data streams, it is entirely possible that some malformed records in that raw data could cause chaos. :thinking: The upside is that they can fix server side bugs without deploying a player client update. :sunglasses:

1 Like
12

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.