4.5 CTD's

[Ray Proudfoot 8,927]

To speed up testing I positioned the Cub at 414 and then switched to the PMDG738 which loads up with engines running. A short taxi to 27L and I took off, flew a circuit and landed back on 27L.

No crash on the original and again after saving when lined up. No RC4 or Ai Companion running. I shall try again after lunch and introduce Ai Companion.

v4.4 with 280 Ai in an 80nm bubble.

[Drumcode 1,892]

4 hours ago, Ray Proudfoot said:

Yes, I have the military option selected. Do you suspect that?

Check Event Viewer. Was there another error a few seconds before the FSUIPC one?

At this point everything is a suspect. 🙂 including ActiveSky and yes, there was another error leading to the FSUIPC one, I'm just not home now to report what it was, but if I remember it was Prepar3d.exe and there was not much to read from it. 

I am going to remove MAIW and all AIG BGLs tonight and try the same flight (ORD-DEN) at the same time of the evening.

[Ray Proudfoot 8,927]

I'm getting really frustrated. Four times I have departed EGLL and flown a circuit and landed and four times there has been no crash. I even ran Ai Companion on the last one just in case.

But all clean as a whistle. I'll switch dates back to Tuesday and try again. It could be an Ai that only flies on that date.

If it was Prepar3D.exe it was likely to be KERNELBASE or something like that.

[Drumcode 1,892]

Ray, Kaiii @ AIG suggested reduction of traffic percentage, did you try that by any chance? 

[Ray Proudfoot 8,927]

2 hours ago, Drumcode said:

Ray, Kaiii @ AIG suggested reduction of traffic percentage, did you try that by any chance? 

Yes and it appears to solve the problem. This is what I wrote on the AIG forum...

After extensive test flights in and out of EGLL I’m pretty certain I know what the problem is. I flew five circuits off 27L at EGLL landing on the same runway. I had EFB running on all and Ai Companion on the last. No crashes at all. Ai was 100%.

So I decided to have a proper flight and loaded a plan into Radar Contact - an ATC addon released back in 2006. It runs on a separate PC via WideFS. Ai % in P3D was set to 100%. As I lined up on the same runway and received my take-off clearance I opened the throttles. Within a couple of seconds P3D had frozen and soon after crashed. KERNELBASE.dll was the cause.

I reloaded everything and tried again but crucially this time I reduced Ai to 50%. No problems at all this time. So the cause is too much Ai being fed into the FSUIPC tables for RC4 and that then causing P3D to crash. The author of FSUIPC is going to install AIG Ai and Radar Contact when time permits and see if he can trap the error and circumvent it. RC4 carries on running so that isn’t crashing. It looks like it’s to do with the amount of data in FSUIPC.

For now I’ll try reduced Ai settings. I’m pleased it isn’t anything in AIG models which I love. Panic over! I’ll revert to v4.5 tomorrow.

[Gandalf 45]

That’s interesting, because I’ve been flying with traffic at 75%... I’ll try reducing it further and hope that does the trick 🙂

Well done Ray for persevering with your testing 👏

Regards

G

[Drumcode 1,892]

Excellent find Ray, once again - you're DA man! 

[Ray Proudfoot 8,927]

Thanks chaps. Am I right in thinking you have only had the crashes when RC4 was running? I cannot recall any flight crashing when it wasn’t running.

Also, it only happens at busy airports. EGLL for me and once at EBBR. Never at EGCC or smaller hubs.

[Gandalf 45]

Happens to me with RC4 running but usually at cruise altitude. My test flight starts at egkk and I don’t get the crash at the airport.

But I will try with reduced AI and see what happens....

[Ray Proudfoot 8,927]

56 minutes ago, Gandalf said:

Happens to me with RC4 running but usually at cruise altitude. My test flight starts at egkk and I don’t get the crash at the airport.

But I will try with reduced AI and see what happens....

Try EGLL. Guaranteed crash when on the runway but never before. It might help if you can monitor the number of Ai in the sim and display it in the Title bar. Let me know if you're interested and I'll supply the offset.

Edited May 31, 2019 by Ray Proudfoot

[Drumcode 1,892]

So what's the safe percentage? 50?

[Gandalf 45]

Drumcode, that’s what I’m going to try this evening.

Ray, that would be useful, thanks

Haven’t commented on thread on the AIG forum but I’ll be sticking with RC4 unless it’s an absolute no go 😀

Edited May 31, 2019 by Gandalf

[Ray Proudfoot 8,927]

@Gandalf, in FSUIPC select the logging tab. On the right there are a few rows. On the top one enter 02A2 and next to it choose U16. Lower down choose Windows Title Bar to see the current number of Ai injected into P3D.

Each to their choice but a couple of posts on AIG did annoy me. RC will continue for as long as we have FSUIPC.

Edited May 31, 2019 by Ray Proudfoot

[Gandalf 45]

Hi All

Just a quick update. I tried with AI set to 50% on my usual scenario and had the same crash with the same 3 faults. Reduced to 40% with the same scenario and completed the flight without a crash.

I did set the ipc logging on in the second test at the approx time of the previous crash. I’ll retry at 50% with the logging on and see what happens, although, as Pete mentions, the files are rather large!

I’ll carry on testing, it just takes a while to keep repeating the same tests and I don’t want to use accelerated time.

Regards

G

[Ray Proudfoot 8,927]

After multiple tests and sending the results to Pete and John Dowson we appear to have found a way of avoiding the CTDs when running RC4 but crucially only when running RC4 via WideFS. This was tested with Ai at 100%.

In the wideclient.ini file locate the [User] section and add this line:-

Deny=2900,12

For those running RC4 on the same PC as P3D a solution has yet to be found. The problem appears to be related to RC4 trying to slew multiple aircraft all at once even those not in any danger of colliding with the user's aircraft. The above command prevents that happening. All the requests come at once causing a problem with P3D. The best way of avoiding the crashes it to limit Ai% to less than 50.

As that code is in RC4 and John Dekker no longer supports the program why that is done will never be known. Hopefully a fix can be found but after 6 hours of testing I'm knackered and there's a certain football match I want to watch this evening.