OOM HELP !

[NBVA330 0]

Good evening people,

The problem is that my P3D V3.3.5 dosen´t start because a notification appears, saying that my computer has run out of memory so that I have to restart the sim.

 

I already try deleting the shaders, re-installing the client and nothing...

MY PC: 
AMD 6300FX

NVDIA GEFORCE 970 GTX

12RAM 

1TB (70GB FREE)

 

HELP !
 

[DaveCT2003 1,820]

Tried deleting your Prepared.CFG? (P3D will rebuilt on the next launch.)

 

If you're running REX Textures at 4096, try running them at 2048 (basically all the settings down one notch).

 

Best wishes my friend.

[vgbaron 3,590]

If you are running a 1TB drive ( I assume a HDD not SSD) and you only have 70G remaining - you have a serious problem. Any HDD that gets over 50% full starts to degrade performance. SSD's can get to 90% before that happens.

 

It would help to know the exact error message.

 

Dave's suggestion is a good one -  give it a try.

 

Vic

[NBVA330 0]

Well at the moment, nothing... I delete the .cfg but dosen´t work...

 

here is a picture 

HELP !

[Poppet 703]

I delete the .cfg but dosen´t work...

 

You deleted the Prepar3D.cfg in  AppData ► Roaming ► Lockheed Martin  ► Prepar3D   Yes  ? 

I see you have the Migration Tool,   Make sure you Disable it  Then Delete the Prepar3D.cfg 

[Luke 782]

 

 

If you are running a 1TB drive ( I assume a HDD not SSD) and you only have 70G remaining - you have a serious problem. Any HDD that gets over 50% full starts to degrade performance. SSD's can get to 90% before that happens.

 

Where do these fascinating statements come from?

 

Hard drives don't care how full they are. You'll get a bit more fragmentation since defrag needs free space to work, but with 70G free that's still plenty. From a performance standpoint you'll have slightly higher latency as the heads have to cross more of the drive, but probably a little more sequential read/write bandwidth since the outside of the drive rotates quicker than the inside. A wash either way.

 

My bet is that the default flight has gotten corrupted, or there's bad scenery/weather if he gets an OOM immediately on startup.

 

Cheers!
 

Luke

[vgbaron 3,590]

Where do these fascinating statements come from?

 

Hard drives don't care how full they are. You'll get a bit more fragmentation since defrag needs free space to work, but with 70G free that's still plenty. From a performance standpoint you'll have slightly higher latency as the heads have to cross more of the drive, but probably a little more sequential read/write bandwidth since the outside of the drive rotates quicker than the inside. A wash either way.

 

My bet is that the default flight has gotten corrupted, or there's bad scenery/weather if he gets an OOM immediately on startup.

 

Cheers!

 

Luke

 

They come from various HDD manufacturers and from other people like NickN and a few other hardware guru's. You are welcome to browse Google if you wish. I stand by my statement.

 

It wouldn't cause his OOM's however.

 

However, I do agree that it's probably a corrupt file and I still recommend he do some serious disk cleanup or get a bigger drive.

 

Vic

[Luke 782]

 

 

They come from various HDD manufacturers and from other people like NickN and a few other hardware guru's. You are welcome to browse Google if you wish. I stand by my statement.

 

I did.

 

here's link 1: http://www.howtogeek.com/215413/why-does-emptying-disk-space-speed-up-computers/

 

And link 2: http://serverfault.com/questions/196956/does-hard-drive-space-affect-performance

 

If you have a citation, I'd love to see it.

 

Cheers!

 

Luke

[vgbaron 3,590]

Your own citations make my point - Take just one factor - fragmentation - more fragmentation when disk is full - more fragmentation equates to slower read times - slower read times equates to degrading performance - which is what I stated.  Thanx for making my point.

 

Not to mention when a drive has less than 10-15% free space, fragmentation may not work ( yes, I know - another quaint notion).

 

Vic

[jabloomf1230 2,445]

A couple of points. As an aside (just in case that anyone who took HS geometry was wondering about it), there is no difference in access/read/write time between files stored on the inner and outer regions of the HDD platter. In the "old days" that might have been true, because of differences in data density, but these days the drive's controller compensates for this disparity.

 

That only leaves file fragmentation as a source of performance loss. As the disk fills up, not only are the individual fragments of big files spread all across the HDD, the remaining free space is also spread out in a similar manner. This makes it more and more difficult for the defrag program to do its job properly. Some 3rd party defrag apps (for example, Raxco PerfectDisk) tend to check to see how often a file is used and then groups all the old files together. This approach leaves a larger block of free space. Basically though, when a HDD gets near to being full, it becomes almost impossible for a defrag app to make any major improvements. 

[vgbaron 3,590]

Agreed - now let's get back to helping the OP solve his issue.

 

Vic

[SteveW 1,792]

You could try rename your dll.xml for a test with the addon .dlls disabled temporarily.

 

These more recent big HDDs have virtual FATs and they shuffle data around as they like it. Defragging these recent drives would have been mostly a waste of time. SSDs try to avoid re-writing the same locations over and over since they can not be re-written to as often as an HDD location. So when small changes happen in files these are created as many fragments purposely.

[Jim Young 2,356]

Agree Steve as the dll.xml is loaded at startup where the crash is occurring.  I've never seen a case where it crashes at startup.  Hopefully he did not place the uiautomationcore.dll in the main P3D folder although I heard PMDG is recommending it.  If he did, he might have put in the wrong one and that module handles memory management for the application.

[Luke 782]

 

 

Your own citations make my point - Take just one factor - fragmentation - more fragmentation when disk is full - more fragmentation equates to slower read times - slower read times equates to degrading performance - which is what I stated.  Thanx for making my point.

 

Fragmented HDDs are slower than non-fragmented HDDs. That is completely orthogonal to free space. If you're fragmented, defrag. Problem solved.

 

 

 

Not to mention when a drive has less than 10-15% free space, fragmentation may not work

 

It has nothing to do with the proportion of free space, and everything to do with the amount of free space relative to the average size of files on the filesystems. He's got 70GB free. Files have gotten bigger, but filesystems have grown even more.

 

As a former colleague of mine used to say, this is a disagreement that can be easily solved with data. Is there a benchmark out there?

 

Cheers!

 

Luke