These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Issues, Workarounds & Localization

 
  • Topic is locked indefinitely.
 

EVE Client freeze/crash issues - Retribution 1.1

First post
Author
MainDrain
The Scope
Gallente Federation
#41 - 2013-02-25 21:31:13 UTC
CCP Mankiller wrote:
We are actively investigating this issue, but our biggest problem is that we cannot reproduce this freeze in-house at all...

We have an idea what might be going on there, but I need your help to prove it:

So the next time you get this freeze, can you please leave it for a few minutes and then open the Windows Task Manager, go to the "Processes" tab and look for a process "ExeFile.exe *32"? When you have found it, keep an eye on the memory consumption. What I'd like to know is: does it go up and up and up? And does it stop finally at the ~4GB boundary?

IF "exefile.exe" has reached this boundary I would like to get a screenshot of your task manager and the client log files! That would help us a lot!









Does it help that the only time ive had this is when ive actually been in a fleet? A large number of disconnects and freezes seemed to occured at the same point in my incursion fleet on saturday and sunday. (couldnt Bug report at the time due to actually trying to log in to save my ship!)
salesman844
Native Freshfood
Minmatar Republic
#42 - 2013-02-25 22:44:03 UTC  |  Edited by: salesman844
I can confirm this as well on my and friends behalf:

My setup:
Operating system: Win 7 Ultimate 64bit
Graphics Card: AMD Radeon HD7850
Driver Version: Catalyst 13.1
CPU: i7-3770
Memory: 16gb
Clients: 4

Friends setup is the same but different socket intel CPU and Nvidia GPU.
No overlays or any unusual software. Never crashed before ret 1.1.1 (or maybe 1.1.2) and now crashes at the rate 1 crash/freeze per hour per computer.
Will try to log but i have to say that when my triage carrier crashes its not logs or screenshots i'm thinking about.

P.S. all the crashes happened in the fleet as well.
Kallehd
Deep Core Mining Inc.
Caldari State
#43 - 2013-02-25 22:48:02 UTC  |  Edited by: Kallehd
Need to know if a single log server will log the entries from both clients? I had a crash tonight, but not from the client I started with the logserver.

If not, you guys need to tell us how to get the log server running for each of the clients as this is like a lottery, where you log one client and the other crashes. Takes too long and I'd like to get this solved faster instead of slower :) Thanks

CCP Vertex / Mankiller ?


EDIT: In case the log server saves data from all clients started after it, I saved the workspace where it happened ... so depending on the answer to the above question, I may or may not have a log server dump from the client freeze.
CCP karkur
C C P
C C P Alliance
#44 - 2013-02-25 23:50:07 UTC
Kallehd wrote:
Need to know if a single log server will log the entries from both clients? I had a crash tonight, but not from the client I started with the logserver.

If not, you guys need to tell us how to get the log server running for each of the clients as this is like a lottery, where you log one client and the other crashes. Takes too long and I'd like to get this solved faster instead of slower :) Thanks

CCP Vertex / Mankiller ?


EDIT: In case the log server saves data from all clients started after it, I saved the workspace where it happened ... so depending on the answer to the above question, I may or may not have a log server dump from the client freeze.

I believe it works for both clients, but CCP Vertex / Mankiller might want to confirm that.

CCP karkur | Programmer | Team Five 0 | @CCP_karkur

The One Neo
The Scope
Gallente Federation
#45 - 2013-02-25 23:58:15 UTC
I have this issue and i'm posting my info here as requested:

CPU: AMD 6-core FX-6100 3.3 GHz / core
GPU: 2 x Ratheon 6670
Drivers: Catalyst version 12.8 (13.1 update caused the need for a windows system restore)
Mem: 8 GB 1600 mhz DDR3
OS: Windows 7 64 bit

my game doesn't just freeze for a few seconds when it does... it freezes until i press the reset button on the computer... i play plenty of other games and this only happens wtih eve

i don't have photobucket and don't see a picture attachment option so i'll list my settings

Display:
Fullscreen
Resolution 1376x768
UI Scaling 100%
interval: intermediate

(skipping to graphic content settings)

Anti-aliasing: disabled
All others: High
Tetania
Republic University
Minmatar Republic
#46 - 2013-02-26 00:01:51 UTC
Sent bug report #154289 no crash dump tho so not sure how much help it will be in narrowing this down. I'll try and remember to start eve with logserv in future.
Shuin Pa
Mentally Imbalanced
#47 - 2013-02-26 04:04:59 UTC
Original post updated with requested screen shots of the task manager.
Jerculees
Common Minerals
#48 - 2013-02-26 04:28:56 UTC  |  Edited by: Jerculees
Jerculees wrote:
I have been getting greyscreen lockup on 2 different computers. Have set up minidump and logserver on one now but haven't played since:

Computer 1
OS: Windows 7 64-bit Professional
GPU: nVidia Geforce GTX 480
Driver Version: 306.97
Number of Clients: 2
RAM: 12GB
Storage drives: Intel 160GB X25-M SSD
CPU: Intel i7-920

One or the other account greyscreens about once per 30 minutes, but stays online for at least 1 hour so far after greyscreen. One account is a freighter pulling ore from an Orca fleet hangar. The second is just a mining boost Orca sitting stationary in an asteroid field. Fleet hangar is used by mining ships to store ore while Freighter removes ore. The Orca stays available and providing boosts and fleet hangar even during greyscreen.


Computer 2

OS: Windows 7 64-bit Ultimate
GPU: nVidia Geforce GTX 680 x 2 (non-SLI)
Driver Version: 314.07
Number of Clients: 5
RAM: 32GB
Storage drives: Samsung 840 120GB SSD (boot), Mushkin Chronos 240GB SSD x 3 in RAID0 (game install)
CPU: Intel i7-3770k

One account runs on 27" monitor using 680-1 card, four accounts run on 24" monitor using 680-2 card. Various of the four accounts greyscreen once per 25 minutes. These are mining ships that continue to mine and stay logged on even after greyscreen. Only 1 account seems to greyscreen at a time so far.

I will provide updates from logserver and minidump as more crashes occur.



@CCP Whoever, if you want to see what a greyscreen lockup looks like:

1) Start up an eve client.
2) Start the snipping tool built into Windows 7
3) Do a NEW rectangular snip.

The screen will haze over white/gray and everything will stop moving but is still running in the background. That is exactly how this greyscreen freeze looks when it happens, except the titlebar will say (Not Responding).

Ok, I played for 2 hours today to monitor CPU, memory and temperatures. I have nothing except EVE running on the computers. No overlays or other extra programs or helpers. After bootup, computer showed cpu-1%, memory-12%, cpu temp-28c. Started 5 eve clients, fleeted up and went into space to mine...cpu-3%, memory-21%, cpu temp-30c, avg exefile *32 memory-540,000K. After 1 hour mining no lockups cpu-10%, memory-22%, cpu temp-31c, avg exefile *32 memory-657,000K. After 2 hours 1 cliented locked up and left running for 20 mins cpu-19% (1 or 2% for 4 working clients, 13% for the locked client), memory-23%, cpu temp-32c, avg exefile *32 memory 640,000 K (for the 4 working clients) / 690,000 K (for the 1 locked client).

On the second computer with 2 clients, 1 locked up with 800,696 K memory constant after lock and 13% cpu while the working client was 722,000 K and 1% cpu. Initial memory for both was 683,000 K after the first hour and then about 715,000 K until the 1 client locked.

There is no crash dump and even the log files seem to show that the client was manually closed by the user ONLY when you restart into a different client session. Tl/dr, it looks like the user interface dies while the rest of the program continues to function in the background.
Rain6637
Simulacra and Simulation
Dracarys.
#49 - 2013-02-26 06:19:15 UTC  |  Edited by: Rain6637
I'm stable currently, and CQ is not hanging tonight.

Task Manager readout for 5 clients:

in Hangar

then in CQ

back to Hangar

back to CQ

memory usage throttles back down to 700's in Hangar, and up to 1GB in CQ, then back down and back up again as I switch between Hangar and CQ.

//I have made a hardware change: I uninstalled my PCI-E sound card.... Soundblaster Recon 3D... the extender ribbon I have to use is a little unstable, and I think when it was plugged in, the Soundblaster mixing program was causing lags throughout my system.

Now I'm a little curious if we all have a sound card in common.
Taredi Lazair
Hedion University
Amarr Empire
#50 - 2013-02-26 08:48:53 UTC  |  Edited by: Taredi Lazair
Windows XP x64
Amd 965
Ati 7750 Catalyst 13.1
Ati 760
8Gb ddr3
No realtek sound drivers installed
Hdd

Windows 7 x64
Amd 965
Ati 785
Nvidia 9600gt 314.07
8Gb ddr2
Via sound drivers installed
sdd

All clients are Fixed windows,interval one and Sounds/voice off.

Number of Clients:
1-10

How long before the freeze/crash occurs?
10min to 3-4 hours

What are you doing when the freeze occurs?:
Mining
CCP Vertex
C C P
C C P Alliance
#51 - 2013-02-26 11:50:04 UTC
Kallehd wrote:
Need to know if a single log server will log the entries from both clients? I had a crash tonight, but not from the client I started with the logserver.

If not, you guys need to tell us how to get the log server running for each of the clients as this is like a lottery, where you log one client and the other crashes. Takes too long and I'd like to get this solved faster instead of slower :) Thanks

CCP Vertex / Mankiller ?


EDIT: In case the log server saves data from all clients started after it, I saved the workspace where it happened ... so depending on the answer to the above question, I may or may not have a log server dump from the client freeze.



Logserver will record all clients running so if you started logserver then started 2 clients it will record both client logs.

**CCP Vertex  |  Senior Development Manager ** | @CCP_Vertex

Freee Lancer
Pod Mods INC
#52 - 2013-02-26 12:05:01 UTC
Hello there

I have the following machines:

Machine 1

Windows 8 x64
Core i7 3770 3.4 ghz
16 GB ram
Geforce GTX 560 TI
SSH hardrive

Machine 2

Windows 8 x64
Core i7 3770 3.4 ghz
AMD Radeon HD 7800
16 GB ram
SSD hard drive

Both these machines pre Retribution could handle anything each by themselves, from mining ops while PVP, etc.

Now that Retribution is out

I have constant client hangs then not responding that i have to force quit in task manager.
this will happen to a few client at random, it does not produce a crash log as when i try close it via the x or the menu it goes white.
i have noticed that the client that hangs is in fact still doing what it is suppose to be for example,
Start mining on 4 clients, after 1 cycle 1 of the clients freezes, i wait till the other 3 are full with cargo then i relog that client that froze and it has a full load.
it is not that they go offline just that the graphic UI stops working.

Secondly
This socket error and all the clients just close.
Please can you explain that
Mexan Caderu
Garoun Investment Bank
Gallente Federation
#53 - 2013-02-26 12:08:12 UTC  |  Edited by: Mexan Caderu
the socket close is a different issue, doesn't have to do with the white client freeze, has more to do with server node and with intrawebs networks.
Freee Lancer
Pod Mods INC
#54 - 2013-02-26 12:12:03 UTC
Mexan Caderu wrote:
the socket close is a different issue, doesn't have to do with the white client freeze, has more to do with server node and with intrawebs networks.


true but,

I am having socket closed, then client freeze, then socket closed, client freeze, client freeze.

and it is starting to get me red in the face, i cant fleet up, run Null/low as i cant have constant connection.

whats the point?
Bubba Hightower
Republic Military School
Minmatar Republic
#55 - 2013-02-26 14:37:14 UTC
Windows 8 64 bit
Intel it-2500k 4 cores 3.3 ghz
16 gigs of Ram
Nvidia GTX 680
tried on two different hard drives, and a reinstall

Ever since Retribution this has happened. Like others have said the client goes grey and says not responding, but if I look in from other clients the individual is still doing his job. I ran 8 clients, went up to 10 recently, every single client is at rock bottom graphical settings in the lowest possible resolution, spending 99% of their time minimized (can you guess what I do :P). Anyway, this happens when they are minimized, using up about 1.5% of CPU per client, with no more than 500 mb of RAM per client, and it happens completely at random. Like I said, if it is my Orca that did this, I could leave it alone and I'm sure it would be boosting all day long, I just can't do anything with the client save force exit. I will check in on the memory usage next time it happens though.

Prior to the big update this has never happened even when playing for long periods of time.
Mexan Caderu
Garoun Investment Bank
Gallente Federation
#56 - 2013-02-26 16:29:23 UTC  |  Edited by: Mexan Caderu
Managed to get one of them white lockups on the logserver, i think...well..the logserver was running when it happened in any case ...

The logserver spat out 2 files, one for each client I guess, one whitescreened, one didn't, I plugged them both in:
http://www.mediafire.com/?3tjgb164np1yulj
http://www.mediafire.com/?gfnz57zam7iii3q

Hope this helps.

Edit: there are unconfirmed rumors that disabling "Smooth camera movement" helps diminish or eliminates the white-outs.
Edit 2, nop, still white screened, with the following windows message (which I'm sure serves no purpose outside microsoft laboratories):
Problem signature:
Problem Event Name: AppHangB1
Application Name: ExeFile.exe
Application Version: 2013.2.48.3842
Application Timestamp: 511b6a0c
Hang Signature: be93
Hang Type: 0
OS Version: 6.1.7601.2.1.0.256.1
Locale ID: 1033
Additional Hang Signature 1: be93684769fb7dcad06c72aa4006aa1a
Additional Hang Signature 2: 8764
Additional Hang Signature 3: 8764fb10f57d11c3f894ce19d2dee5ce
Additional Hang Signature 4: be93
Additional Hang Signature 5: be93684769fb7dcad06c72aa4006aa1a
Additional Hang Signature 6: 8764
Additional Hang Signature 7: 8764fb10f57d11c3f894ce19d2dee5ce


Mexan Caderu wrote:
Operating System: win7 64 bit
Graphics Card: nvidia gtx 550ti
Driver Version: appHanged on both 310.90 and 314.07
Number of Clients: usually 2, of which only 1 hanged so far, probably because the client/char that doesn't freeze is in station or just sitting in space idly
How long before the freeze/crash occurs? usually between 15 minutes to 5 hours
What are you doing when the freeze occurs?: Incursion running (explosions,fleet activity, lasers, turret effects ?). Never hanged during other activity.

screenshot with graphical settings: http://i.imgur.com/O60UjjF.png -> has also hanged with max settings and every graphical effect ticked on too.
Eve dump: not available. Nothin' is created in the specified location, perhaps because eve doesn't think it crashed.


Freeze symptoms:
* Logged in, business as usual, etc, etc
* then suddenly windows cursor goes into 'busy'
* clicking the eve window makes the eve window get a gray-white overlay
* EvE window is "not responsive". Windows offers to close the exefile.exe program that is not responding.
* task manager shows the single exefile.exe is saturating 1 core of the cpu (100% utilization of that core, normally core utilization is 2-10%)
* During this, the char is still in-game, alive and kicking
* alt-F4 or close the program as per windows offer
* only now reports come over coms that the char has DC'ed

CCP Vertex wrote:
We have had a small number of reports of client lockup/freezes and sometimes crashes.

I believe the freezes are rather frequent insome circumstances. In the incursion HQ fleet I tend to frequent there is usually one "white lockup" every 5-10 minutes.
Kallehd
Deep Core Mining Inc.
Caldari State
#57 - 2013-02-26 18:47:32 UTC
CCP Vertex wrote:
Kallehd wrote:
Need to know if a single log server will log the entries from both clients? I had a crash tonight, but not from the client I started with the logserver.

If not, you guys need to tell us how to get the log server running for each of the clients as this is like a lottery, where you log one client and the other crashes. Takes too long and I'd like to get this solved faster instead of slower :) Thanks

CCP Vertex / Mankiller ?


EDIT: In case the log server saves data from all clients started after it, I saved the workspace where it happened ... so depending on the answer to the above question, I may or may not have a log server dump from the client freeze.



Logserver will record all clients running so if you started logserver then started 2 clients it will record both client logs.



CCP Vertex, I submitted a Bug Report from the crash I had last night. Number is 154336.

Please let me know if you need more log server dumps to solve this as it's getting really annoying.
Shuin Pa
Mentally Imbalanced
#58 - 2013-02-27 01:56:31 UTC  |  Edited by: Shuin Pa
Client will not white out with logserver running... The bug knows it's running and refuses to rear its ugly head...

The machines are smarter than us.

BTW tested the fleet theory, I think it is coincidence. Currently in fleet running the log server so I predict it will not whiteout at all and everything will be fine until I stop the log server and go to PVP. 8-p

Signed,


Shuin "Frustrated" Pa


E D I T!! --> That was classic! no sooner than I posted this message did the client crash LMAO

Bug Report 154357

I hope I uploaded the correct file. This time my SD client crashed, and yes, while in fleet.
If you need the other log server files in that dir tell me how to get them to you
MainDrain
The Scope
Gallente Federation
#59 - 2013-02-27 10:12:21 UTC
RE the sound issue, although not a crash, i figured it could go here.

I have noticed that if i lock targets on a ship and he decides he wants to dock up I get an annoying popping sound when i zoom in and out, with the sound being cut off, a quick fix is to disable and renable audio, but not ideal
Rain6637
Simulacra and Simulation
Dracarys.
#60 - 2013-02-27 10:20:34 UTC  |  Edited by: Rain6637
one of my clients jumped to 10MB/sec disk transfer speed, and I'm not 100% sure if it was the only one... the others could have done the same, and I just missed it.

dual intel 330 series 120GB SSDs in stripe (RAID 0)

cpu, RAM, disk use, etc in CQ and in space. (higher readings are in CQ)