Here's my findings so far, Test 1.

 

1) System: Windows 7 64 Professional, Prepar3D 2.5.12945.0 (Full Installer, not hotfix). PMDG 777 base and expansion installed 1.10.6492. Process Explorer used to check VAS used.

 

2) P3D Settings/conditions: No weather, no AI, sat at gate 536 (T5) Aerosoft EGLL, FTX Global, Vector, OpenLC installed. Time of day 11am sim time. Loaded EGLLVHHH flightplan created by PFPX including VHHH arrival SIERRA1G and approach ILS 25R.

 

3) Test: I sat at the gate in the VC without touching anything after loading the flightplan except to check the VAS usage. VAS started at 3.222GB "Virtual Size" in process explorer. 3 hours later VAS had risen to 3.516GB without me touching anything or changing view. 6 hours after beginning test VAS had risen further to 3.844GB.

 

 

Test 2

 

1) As test 1

 

2) As test 1 with the exception that I didn't load an arrival or an approach into the FMC. Everything else was identical.

 

3) As test 1 with the exception that VAS started at 3.230GB "Virtual Size" in process explorer. 3 hours later VAS had decreased slightly to 3.223GB. 6 hours after beginning VAS decreased to  3.222GB.

 

So there seems to be a definite correlation between VAS leakage and loading an arrival and approach into the FMC at the departure airport.

 

My next test will be to execute identical flights with and without approach/arrival info loaded and see how far I get with each. Sim conditions will be identical to tests 1 & 2. Will post back my results.

I'll run without Active Sky

Also, the weather needs to be set to Clear Skies in FSX.

Glad to see someone else is experiencing this issue still. I posted another thread yesterday of a similar situation that resulted in an OOM when the sim was running for a much longer period with the WX radar on vs running a couple hours with the WX radar off. In both situations I loaded an ILS approach early on in the flight.

Deleted by user.

Glad to see someone else is experiencing this issue still. I posted another thread yesterday of a similar situation that resulted in an OOM when the sim was running for a much longer period with the WX radar on vs running a couple hours with the WX radar off. In both situations I loaded an ILS approach early on in the flight.

 

Link that contains details, please?

Link that contains details, please?

 

http://forum.avsim.net/topic/468866-oom-at-cruise/

http://forum.avsim.net/topic/468866-oom-at-cruise/

 

Thanks.

 

 

All, I'm going to try and assemble a controlled scenario for pilots to try and post here, to see if I can extract some results that are at least in a semi-consistent baseline.  Then we can start adding, slowly, some variables.  Since there seems to be a shotgun smattering of data in several threads the past few days, I'll see if I can try to do some data reduction.

 

(Bear with me though, flight testing real aircraft for a living is eating up most of my free time this month and next.  It's a shame those real airplanes get in the way of my fun with virtual ones .  Also, unfortunately, I only own the 77L, not 77W, so I can't reproduce those specific test points myself, but hopefully I can at least try to identify some trends, if there are any.)

Thanks.

 

 

All, I'm going to try and assemble a controlled scenario for pilots to try and post here, to see if I can extract some results that are at least in a semi-consistent baseline.  Then we can start adding, slowly, some variables.  Since there seems to be a shotgun smattering of data in several threads the past few days, I'll see if I can try to do some data reduction.

 

(Bear with me though, flight testing real aircraft for a living is eating up most of my free time this month and next.  It's a shame those real airplanes get in the way of my fun with virtual ones .  Also, unfortunately, I only own the 77L, not 77W, so I can't reproduce those specific test points myself, but hopefully I can at least try to identify some trends, if there are any.)

 

Excellent, I'd love to be a part of the process to help narrow down the issue.

 

 

So there seems to be a definite correlation between VAS leakage and loading an arrival and approach into the FMC at the departure airport.

 

I did a short comparison today with similar conditions, but my findings are inconclusive and point to a need for a pair of longer test runs.  My results were 243MB increase in VAS with the VHHH arrival and approach and 212MB increase without the VHHH arrival and approach.  Test conditions are:  Flight is created at the gate on ground air and electrical and saved.  Both trials are run with loading that saved flight and going only as far as setting up the route in the FMS (FMS is set up but no APU, pushback or engine start). Same EGLL as OP but I'm in an AAL B77W at gate 343, no weather, no traffic, etc.  Only difference between trials is one with and the other without entering the SIERRA1G arrival and TD.ILS25R approach.  Both trials had an increase in VAS use (by an amount I consider nominal) with a difference in increase of only 23MB.  During beta testing current B77X version, I found that a difference in VAS of 200MB to be normal between any two trials of same test conditions.

 

I believe the test conditions should include actual flight time of at least 6 hours, which means testing is going to take time.

Here's my findings so far, Test 1.

 

1) System: Windows 7 64 Professional, Prepar3D 2.5.12945.0 (Full Installer, not hotfix). PMDG 777 base and expansion installed 1.10.6492. Process Explorer used to check VAS used.

 

2) P3D Settings/conditions: No weather, no AI, sat at gate 536 (T5) Aerosoft EGLL, FTX Global, Vector, OpenLC installed. Time of day 11am sim time. Loaded EGLLVHHH flightplan created by PFPX including VHHH arrival SIERRA1G and approach ILS 25R.

 

3) Test: I sat at the gate in the VC without touching anything after loading the flightplan except to check the VAS usage. VAS started at 3.222GB "Virtual Size" in process explorer. 3 hours later VAS had risen to 3.516GB without me touching anything or changing view. 6 hours after beginning test VAS had risen further to 3.844GB.

 

 

Test 2

 

1) As test 1

 

2) As test 1 with the exception that I didn't load an arrival or an approach into the FMC. Everything else was identical.

 

3) As test 1 with the exception that VAS started at 3.230GB "Virtual Size" in process explorer. 3 hours later VAS had decreased slightly to 3.223GB. 6 hours after beginning VAS decreased to  3.222GB.

 

So there seems to be a definite correlation between VAS leakage and loading an arrival and approach into the FMC at the departure airport.

 

My next test will be to execute identical flights with and without approach/arrival info loaded and see how far I get with each. Sim conditions will be identical to tests 1 & 2. Will post back my results.

 

Test 3

 

1) As tests 1 & 2.

 

2) P3D Settings/conditions: No weather, no AI, sat at gate 536 (T5) Aerosoft EGLL, FTX Global, Vector, OpenLC installed. Time of day 11am sim time. Loaded EGLLVHHH flightplan created by PFPX including the SIERRA1G arrival but no approach.

 

3) Test: I flew the flightplan with a starting VAS of 3.018GB used. I flew in real time (no time compression) and 70nm from TOD I loaded the TD.ILS 25R approach. I landed, and taxied to gate W40 (Taxi2Gate VHHH, NO GSX). After cutting engines I checked VAS usage which was 3.782GB after 11+hrs flight time.

 

 

Test 4

 

1) As test 3

 

2) P3D Settings/conditions: No weather, no AI, sat at gate 536 (T5) Aerosoft EGLL, FTX Global, Vector, OpenLC installed. Time of day 11am sim time. Loaded EGLLVHHH flightplan created by PFPX including VHHH arrival SIERRA1G and approach TD.ILS 25R.

 

3) Test: I flew the flightplan with a starting VAS of 3.029GB used. I flew in real time (no time compression) and the ping of death began to ring on approach to TD VOR. I continued to fly the approach and 10nm away from the runway threshold I had an OOM. VAS was just over 4GB before P3D shut down.

 

Conclusion: Loading the approach into the FMC being the only variable, one has to conclude that this lead to an increase in VAS consumption over test 3. (N.B. The same livery was used in all 4 tests).

Test 3

 

1) As tests 1 & 2.

 

2) P3D Settings/conditions: No weather, no AI, sat at gate 536 (T5) Aerosoft EGLL, FTX Global, Vector, OpenLC installed. Time of day 11am sim time. Loaded EGLLVHHH flightplan created by PFPX including the SIERRA1G arrival but no approach.

 

3) Test: I flew the flightplan with a starting VAS of 3.018GB used. I flew in real time (no time compression) and 70nm from TOD I loaded the TD.ILS 25R approach. I landed, and taxied to gate W40 (Taxi2Gate VHHH, NO GSX). After cutting engines I checked VAS usage which was 3.782GB after 11+hrs flight time.

 

 

Test 4

 

1) As test 3

 

2) P3D Settings/conditions: No weather, no AI, sat at gate 536 (T5) Aerosoft EGLL, FTX Global, Vector, OpenLC installed. Time of day 11am sim time. Loaded EGLLVHHH flightplan created by PFPX including VHHH arrival SIERRA1G and approach TD.ILS 25R.

 

3) Test: I flew the flightplan with a starting VAS of 3.029GB used. I flew in real time (no time compression) and the ping of death began to ring on approach to TD VOR. I continued to fly the approach and 10nm away from the runway threshold I had an OOM. VAS was just over 4GB before P3D shut down.

 

Conclusion: Loading the approach into the FMC being the only variable, one has to conclude that this lead to an increase in VAS consumption over test 3. (N.B. The same livery was used in all 4 tests).

I encourage you also to participate in this thread to capture and baseline VAS data: http://forum.avsim.net/topic/468925-the-pmdg-777-vas-data-collection-thread/

 

 

Loaded EGLLVHHH flightplan created by PFPX including VHHH arrival SIERRA1G and approach TD.ILS 25R.

 

This is an assumption that I would like to close:  PMDG does not import the arrivals or approach in flightplans not saved by PMDG, so my assumption is after loading the route you then selected the arrival and approach while on the ground during pre-flight?

 

 

This is an assumption that I would like to close:  PMDG does not import the arrivals or approach in flightplans not saved by PMDG, so my assumption is after loading the route you then selected the arrival and approach while on the ground during pre-flight?

 

You're a clever man Dan!

Excellent powers of deduction there buddy.

Installed the -300 on my machine over the weekend and flew this exact route (except used ILS 28C until 40 NM out, then switched to ILS 28R since my scenery doesn't actually have 28C at KORD).  Did not fly online and used ActiveSky Next in lieu of ActiveSky Evolution.  Also flew UK2000 EGLL v3 (with normal scenery amount selected in the configuration utility) and FSDT KORD.  Used normal autogen, 0% AI, and DX10.  Also used FTX Global but not Vector.  Preselected the approach at the gate in EGLL and have not been able to reproduce this leak report since the latest update on the -200 or -300.

 

VAS at initial load: 2.71 GB

VAS at pushback: 2.80 GB

VAS at TOC: 2.62 GB

VAS at TOD: 2.65 GB

VAS at taxi in: 3.00 GB

Agree Kyle.  There's no reason why the 77W would behave differently than the 77L in this regard and I've flown a couple of 77W flights since this post where there was no indication of a problem other than what is normally encountered with the VAS limit.