P3Dv4 Pause (Not Responding) Then Back

[simbol 5,814]

1 hour ago, Cruachan said:

Hi Dan,

If true then it's a heck of a long frame! However, I do note that Core 0 is running at 100% during the pause while the 1080Ti is taking a short break! Of course, this raises the question as to what on earth could be responsible for pushing the overclocked 5960X @ 4.78GHz in this relatively unchallenging situation?!

I run with HT=ON and an Affinity Mask value of 21845 in Prepar3D.cfg, so 'only' 8 cores active.

Why are others not confirming this at KVPS/HRT? To be fair, other than my friend with his virgin 4.1 installation and a far less potent rig, nobody here has chipped in. So, it's difficult to draw any conclusions.

Regards,

Mike

What happens if you change the affinity settings to something different? I am just wondering if it would any impact.

Simbol 

[DylanM 1,069]

^ or remove it completely?

[downscc 1,327]

My configuration works best with no AM specified.  Also, looking at core activity doesn't tell you much, the processor could be maxed out just doing a look-wait-look loop and not actually doing anything while the process waits for another process to finish.  Be careful when interpreting CPU loading, you cannot compare it to getting work done.

[Cruachan 705]

3 hours ago, simbol said:

What happens if you change the affinity settings to something different? I am just wondering if it would any impact.

Simbol 

Hi Simbol, DylanM

In fact I decided to try the effect of disabling the Affinity Mask (16 LPs active).

Result: essentially very little difference in behaviour:

Screen captured during the pause....

First large downward spike represents a brief hesitancy (long frame) while rolling down rwy 19 for takeoff.

The second spike represents the extended (4-5 second) pause as seen before with the Affinity Mask enabled (8 cores).

Note: GPU usage trace appears 'cleaner' with less variance using 8 cores rather than with 16 LPs active:

Regards,

Mike

[simbol 5,814]

4 minutes ago, Cruachan said:

Hi Simbol, DylanM

In fact I tried the effect of disabling the Affinity Mask (16 LPs active).

Result: essentially very little difference in behaviour:

Mike

Ok, it must be something we are missing, I am just trying to fit the puzzle, I wonder if people experiencing this problem has a similar motherboard or chipset.

Have you upgraded your bios firmware to the latest version?

I was having issues with my EVGA Z170 motherboard due to microcode bugs, this was causing random crashes and problems with some apps (including PD3V4) when my CPU was under load with certain OC speeds, I know is nothing similar to the issue on this topic but perhaps is a factor that we could consider.

Regards 

Simbol 

[Cruachan 705]

34 minutes ago, simbol said:

Ok, it must be something we are missing, I am just trying to fit the puzzle, I wonder if people experiencing this problem has a similar motherboard or chipset.

Have you upgraded your bios firmware to the latest version?

I was having issues with my EVGA Z170 motherboard due to microcode bugs, this was causing random crashes and problems with some apps (including PD3V4) when my CPU was under load with certain OC speeds, I know is nothing similar to the issue on this topic but perhaps is a factor that we could consider.

Regards 

Simbol 

Hi Simbol,

Yes, my BIOS is current. However, with all respect I cannot see how this would be relevant as these pauses are not random in nature. They always occur in the same places (the few places where I have encountered them to date) and appear to be predictable, at least that's my experience. To be frank, their occurrence around the virtual globe is rare* and seem to be confined to specific situations. If I wish to fly in an area like KVPS/HRT then a few flights without resetting the sim seems to get rid of these events and all is well. However, I can see that others would find it irritating to encounter a brief halt while, say, approaching the end of an otherwise trouble-free flight.

What I find difficult to understand is why I'm not experiencing these pauses in far more demanding situations than those presenting at the current Prepar3D v4.1 Default situational area. Also, I should underline the fact that on my rig this phenomenon is restricted to Prepar3D only. You will get a flavour of my other installations from the desktop icons visible in a previous screen capture.

*Rare as in places I have visited so far.

For me it's not a major issue, but since the topic was posted I thought it worth having a bash at identifying the cause,

Regards,

Mike

[simbol 5,814]

Thanks Mike, from what you are describing it seems to be a bug on the content scenary, I gather that this occurs to you even with all the PD3V4 settings set at the lower level  correct?

Best Regards 

S.

[Cruachan 705]

5 hours ago, simbol said:

Thanks Mike, from what you are describing it seems to be a bug on the content scenary, I gather that this occurs to you even with all the PD3V4 settings set at the lower level  correct?

Best Regards 

S.

Okay, I have returned P3D to its default settings by deleting Prepar3D.cfg and allowing a rebuild of the file. Nothing else has changed.

2 Images follow - The first captured during the pause and the second captured after flying around for 7 minutes. This included several approaches and passes over HRT.

During Pause

After FlyAround

So, this extended pause still occurs at lower P3D settings and no similar pauses were encountered during the subsequent 7 minute flight. Once again Core 0 rose to 100% during the pause.

Regards,

Mike

 

[simbol 5,814]

4 minutes ago, Cruachan said:

this extended pause still occurs at lower P3D settings and no similar pauses were encountered during the subsequent 7 minute flight. Once again Core 7 rose to 100% during the pause.

Regards,

Mike

 

Well at least now we know this has nothing to do with the sim settings.

I think you should report this to LM now that you have all the facts and tests so well documented.

Many thanks for all your reports they are invaluable.

Best Regards 

Simbol 

[Cruachan 705]

Thanks Simbol for your helpful input. It's appreciated. 

I don't wish to muddy the waters, but I've reminded myself that the title of this thread by Craig included the words 'Not responding'.

This message has never appeared during my testing so it raises the possibility that our respective pauses have different causal origins and Craig and I may be describing/experiencing separate issues, although the effects appear similar.

At this stage it would be useful to hear about the experiences of some other simmers. I note that the thread view count is fairly significant so it appears it has been attracting some interest.

Best regards,

Mike

[DylanM 1,069]

Hi Mike, as this is a reproducible problem, I'd also suggest posting this on the LM support forum for them to investigate.

[Cruachan 705]

14 minutes ago, DylanM said:

Hi Mike, as this is a reproducible problem, I'd also suggest posting this on the LM support forum for them to investigate.

Hi Dylan,

Do you think it would be okay to simply provide a link to this thread? Are they likely to follow the link? It would certainly save me some time!

Regards,

Mike

[DylanM 1,069]

I'm not sure Mike...but it may be best to copy and paste the text and supporting images into your post there (as the devs themselves will review and respond there).

[jason99vmi 157]

Had this issue occasionally in P3D v3.4 but hadn't seen it in v4 until last night.  Was flying AS CRJ 7 into DD KEWR last night and experienced this on final.  Figured it had to do with the amount of autogen.  NYC area is notorious for this issue.  

Certainly interested though if it could be something else.  

DD KEWR also uses SODE so maybe that was it.  

[mwilk 1,106]

I started noticing this issue a few weeks ago. Specifically, when taking off from KDAL on runway 31L. At the same point every time the sim pauses for 3 or 4 seconds and then resumes. I've seen it happen in other places to. I tried disabling various addons to see if that might have been the cause but nothing worked. Very odd.