Maintenance for the week of November 25:
• PC/Mac: NA and EU megaservers for maintenance – November 25, 4:00AM EST (9:00 UTC) - 7:00AM EST (12:00 UTC)
• Xbox: NA and EU megaservers for maintenance – November 27, 6:00AM EST (11:00 UTC) - 9:00AM EST (14:00 UTC)
• PlayStation®: NA and EU megaservers for maintenance – November 27, 6:00AM EST (11:00 UTC) - 9:00AM EST (14:00 UTC)

TG crashing

  • LaucianNailor
    LaucianNailor
    ✭✭✭
    @KhajitFurTrader
    Would be interesting to know how much video RAM your Mac has, @LaucianNailor. @Mardi Gras's as well. You can see it listed in the Graphics entry of Apple Menu > About this Mac.


    I've the 4GB Nvidia GTX 780M. System RAM is 24GB
    Edited by LaucianNailor on March 9, 2016 7:14PM
    Mac/PC EU Server

    Lots of alts....561+ CP
    Inside Trade Guild
  • KhajitFurTrader
    KhajitFurTrader
    ✭✭✭✭✭
    ✭✭
    Seems the 2013 27" iMac with the GTX 780M is quite popular around here, I have the same machine. :)

    So, from the reports, the amount of video RAM is ample, and lack of it not very likely the cause. Although I concur with your initial theory that seeing bare meshes (should never happen) likely occurs by changing levels of detail, i.e. when low-res textures get exchanged for high-res ones.

    I have yet to see this issue myself.

  • Atlan_daGonozal
    Atlan_daGonozal
    ✭✭✭
    Me too ;-)

    Error on opening /dev/urandom — what could have gone wrong there:
    eso                   
    ****************************
    ESO Start Up! 
    ****************************
    Args = eso  
    ****************************
    Current Working Directory = /Applications/Spiele/The Elder Scrolls Online EU/game_mac/pubplayerclient/eso.app/Contents/Resources/ 
    ****************************
    libc++abi.dylib: terminating with uncaught exception of type CryptoPP::OS_RNG_Err: OS_Rng: open /dev/urandom operation failed with error 24
    

  • KhajitFurTrader
    KhajitFurTrader
    ✭✭✭✭✭
    ✭✭
    Me too ;-)

    Error on opening /dev/urandom — what could have gone wrong there:
    Have you tried the workaround from this thread: 2.3.5 64-bit client FAQ (4th section)?

    Edited by KhajitFurTrader on March 9, 2016 8:30PM
  • Atlan_daGonozal
    Atlan_daGonozal
    ✭✭✭
    Have you tried the workaround from this thread: 2.3.5 64-bit client FAQ (4th section)?

    Yes, just found. Still crashes — :'( — but the error message is longer :| :
    >eso            
    ****************************
    ESO Start Up! 
    ****************************
    Args = eso  
    ****************************
    Current Working Directory = /Applications/Spiele/The Elder Scrolls Online EU/game_mac/pubplayerclient/eso.app/Contents/Resources/ 
    ****************************
    Launching Crash Handler:
      EXE : /Applications/Spiele/The Elder Scrolls Online EU/game_mac/pubplayerclient/eso.app/Contents/Resources/ZoCrashReporter.app/Contents/MacOS/ZoCrashReporter
      Args: en \\.\pipe\ZoCrashReport\eso\28986a84 "Local\ZoCrashReport.SyncEvent.eso.28986a84" "/Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/"
    arg 2, Arg == en 
    arg 3, Arg == \\.\pipe\ZoCrashReport\eso\28986a84 
    arg 4, Arg == Local\ZoCrashReport.SyncEvent.eso.28986a84 
    arg 5, Arg == /Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/ 
      PID: ab47
    Requesting Dump
    Polling RequestDumpForException
    
    -------------------------------------------------------
    ZeniMax Online Studios - Error Reporter
    
    Arg Count (5 args, expected 5)
    Printing Arg 0 
    [Arg 0] /Applications/Spiele/The Elder Scrolls Online EU/game_mac/pubplayerclient/eso.app/Contents/Resources/ZoCrashReporter.app/Contents/MacOS/ZoCrashReporter 
    
    Printing Arg 1 
    [Arg 1] en 
    
    Printing Arg 2 
    [Arg 2] \\.\pipe\ZoCrashReport\eso\28986a84 
    
    Printing Arg 3 
    [Arg 3] Local\ZoCrashReport.SyncEvent.eso.28986a84 
    
    Printing Arg 4 
    [Arg 4] /Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/ 
    
    
    Copying Args to Strings 
    
    Printing Arg Strings 
    Language Code: en
    Pipe Name: \\.\pipe\ZoCrashReport\eso\28986a84
    Event Name: Local\ZoCrashReport.SyncEvent.eso.28986a84
    Crash Report Path: /Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/
    
    g_path: /Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/
    -------------------------------------------------------
    -------------------------------------------------------
    Start: 1
    Saving the details of your error...
    ClientDumpRequestCallback... 
    ClientDumpRequestCallback apDumpFilename.c_str() /Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/65DBA771-34AC-4608-9365-434D32A6CABE.dmp 
    ImportParameters... 
    
    ====================================================================
      e has stopped working.
    ====================================================================
    
    ClientDumpRequestCallback About to unlock 
    ClientDumpRequestCallback Unlock exiting 
    -------------------------------------------------------
    ZoCrashReporter - Saved Folder Path: /Users/martin/Documents/Elder Scrolls Online/liveeu/Errors/
    GetFirstReport.
    GetReport::ForEachResult 
    aEntry.pName: 65DBA771-34AC-4608-9365-434D32A6CABE.dmp
    ZoCrashReporter - Saved File: 65DBA771-34AC-4608-9365-434D32A6CABE.dmp
    WriteExtraFile Path 65DBA771-34AC-4608-9365-434D32A6CABE.extra 
    Waiting on Child Process
    ZoCrashReporter - Crash Dump File Loaded 
    Uploading error reports.
    UploadReport::ForEachResult 
    aEntry.pName: 65DBA771-34AC-4608-9365-434D32A6CABE.dmp
    Uploading Data to URL: http://live-crash-reporter.elderscrollsonline.com/submit
    UploadReport::ForEachResult 
    Child Process Done
    
    
  • LaucianNailor
    LaucianNailor
    ✭✭✭
    Seems the 2013 27" iMac with the GTX 780M is quite popular around here, I have the same machine. :)

    So, from the reports, the amount of video RAM is ample, and lack of it not very likely the cause. Although I concur with your initial theory that seeing bare meshes (should never happen) likely occurs by changing levels of detail, i.e. when low-res textures get exchanged for high-res ones.

    I have yet to see this issue myself.

    Hi KT,
    To update you on this, I've just had a note from Chris who tells me It's apparently a fault in the OGL renderer which is carrying out the LOD operation too quickly. The graphics team are aware of it and looking into a fix but do eta on this as yet.
    Mac/PC EU Server

    Lots of alts....561+ CP
    Inside Trade Guild
  • Pomaikai
    Pomaikai
    ✭✭✭✭✭
    How do you release with a Hotfix with a bug that causes huge numbers of your Mac customers to crash when looting ANYTHING in the game? I think that's pretty obvious, you don't actually test it. Not to be overly rude, but when I'm alpha and beta testing software and hardware, I don't pay for the privilege of doing so. When I am paying for software, it's release ready Gold software. This has to stop. Test before releasing. If you need Mac users to test a special hot fix, then send out invites to your ESO+ Mac users and ask them for help if you don't have the internal staff and alpha test group to do a thorough test. You don't use internal builds that are several versions ahead of the ones your beta testers are currently testing. You release what they have tested and vetted. You don't release untested code into the wild. This is what's currently going on with the game, and it's more than just an annoyance. It's repeated poor development processes.
  • KhajitFurTrader
    KhajitFurTrader
    ✭✭✭✭✭
    ✭✭
    Have you tried the workaround from this thread: 2.3.5 64-bit client FAQ (4th section)?

    Yes, just found. Still crashes — :'( — but the error message is longer :| :
    That's the ZeniMax CrashReporter at work. :smiley:

    Just to make sure: did you reboot your machine recently, or are you like me and let it sleep in-between sessions, so that the OS runs for days or even weeks on end?

    Please try the launchctl solution @smacx250 has posted to the FAQ thread, rebooting once after the change to the maximum file handle number has been made. After logging back in, verify the change like @smacx250 suggested, then launch the client (manually or via the Launcher, should make no difference).

    To update you on this, I've just had a note from Chris who tells me It's apparently a fault in the OGL renderer which is carrying out the LOD operation too quickly. The graphics team are aware of it and looking into a fix but do eta on this as yet.
    Awesome! That pesky bow FPS-drop bug is in their sights as well.
  • smacx250
    smacx250
    ✭✭✭✭✭
    Have you tried the workaround from this thread: 2.3.5 64-bit client FAQ (4th section)?

    Yes, just found. Still crashes — :'( — but the error message is longer :| :
    That's the ZeniMax CrashReporter at work. :smiley:

    Just to make sure: did you reboot your machine recently, or are you like me and let it sleep in-between sessions, so that the OS runs for days or even weeks on end?

    Please try the launchctl solution @smacx250 has posted to the FAQ thread, rebooting once after the change to the maximum file handle number has been made. After logging back in, verify the change like @smacx250 suggested, then launch the client (manually or via the Launcher, should make no difference).

    To update you on this, I've just had a note from Chris who tells me It's apparently a fault in the OGL renderer which is carrying out the LOD operation too quickly. The graphics team are aware of it and looking into a fix but do eta on this as yet.
    Awesome! That pesky bow FPS-drop bug is in their sights as well.
    Actually, I think that rebooting after doing the launchctl command will clear it back to the defaults. Otherwise, any new program run after the launchctl command should run with the new value (anything already running won't get it). To make the change persist through a reboot requires creating a launchctl startup file. If there is interest, I can add that to the post as well.
  • Rosiv
    Rosiv
    Useful, indeed, but still a workaround.

    I would prefer to see a working patch for this.
    There were two hotfixes in the past, but looting is still a mess.
    It's always the same: I log in and first loot crashes the client until I remember to disable auto loot again.
    Changing settings in the OS is not a possibility for me, because I don't know the exact impacts resulting from this.

    At least there IS a workaround discussed here. ;)
  • smacx250
    smacx250
    ✭✭✭✭✭
    Yes, you can certainly choose not to use the maxfiles workaround if you feel uncomfortable about it. I'm sure this issue will be patched, but I don't know that we have heard when that will be. I can give you an idea about what the workaround does.

    In unix (which OS X is), each process (or program) has a limit on how much computer resources it can consume. Since there are many processes, the system needs to make sure that no single process can consume "too much" of the overall. The number of open files is one of those limits. Furthermore, there are two limits - "soft" and "hard". When a process hits the "soft" limit, the system sends the process an indication letting it know that it reached the soft limit. The process can ignore the soft limit if it chooses to do so, or it can take some action. In the case of ESO, it chooses to crash. If a process ignores the soft limit warning and continues to open more files such that it hits the hard limit, it will experience a hard failure - it can not be ignored. From the "getrusage" man page:

    A resource limit is specified as a soft limit and a hard limit. When a soft limit is exceeded a process may receive a signal (for example, if the cpu time or file size is exceeded), but it will be allowed to continue execution until it reaches the hard limit (or modifies its resource limit).

    On 10.11, the default soft and hard limits are 256 and 10240, respectively (described by some as "ridiculously low"). What the change does is increase the soft limit from 256 to 4000. It leaves the hard limit at the default ("unlimited" in this case will actually end up being 10240 due to how the system handles the hard limit). This can be done such that it only affects ESO, in which case the limit must be changed and ESO run from within the domain of that change (what the FAQ describes when running ./eso directly from the Terminal). It can also be done globally, such that all new programs run will get the limit (which is what the "launchctl" version of the workaround does). It may be that you feel more comfortable with the local workaround than the global one - or not at all.
  • KhajitFurTrader
    KhajitFurTrader
    ✭✭✭✭✭
    ✭✭
    That was a great explanation, @smacx250. :smiley:
    smacx250 wrote: »
    I'm sure this issue will be patched, but I don't know that we have heard when that will be.
    Could be as early as next Monday, because a fix has already been checked in, or so I heard. ;)

    In this particular case, it's an easy fix, because all the client process has to do when launching is to tell the OS "Set my (file) handle soft limit to 5000", and it's good to go. :smiley:


    Edited by KhajitFurTrader on March 10, 2016 11:59AM
  • smacx250
    smacx250
    ✭✭✭✭✭
    Thanks for the update, @KhajitFurTrader!

    As an aside, with the latest patch and the maxfiles workaround (and disabling auto-loot for the first loot), last night I experience probably the most stable play I've had in months (yes, pre-TG even!). I played in Cyrodiil for over two hours straight without any crashes. Pre-TG, my client would usually crash in transit after about an hour of play (workaround was to try and remember to quit and restart before it crashed - I often forgot!). The play was better in other ways as well - mounted sprinting from the Orsinium wayshrine to the writ drop location resulted in no loading screens (previously I always hit two), the crafting benches and guild trader NPCs in Rawl'kha showed up much more quickly, and the guild trader NPCs didn't disappear when I ran to the other end of the town and back. Also, FPS was improved. Still some graphical glitches here and there, but in all, starting to look pretty good! I'm hopeful future sessions will also go well!
  • Rosiv
    Rosiv
    Does this

    • European PC/Mac megaserver for patch maintenance – March 15, 2:00 GMT (March 14, 10:00PM EDT)

    prepare the server for the Mac-Patch? ;)
  • Rosiv
    Rosiv
    I know, but we're all knowing here, what we are waiting for! ;)
  • Randay
    Randay
    ✭✭✭
    Rosiv wrote: »
    I know, but we're all knowing here, what we are waiting for! ;)
    Yup!!
    File size handle limit adjustments and auto loot crashing bug fix.
    if less really is more, then maybe nothing is everything
  • KhajitFurTrader
    KhajitFurTrader
    ✭✭✭✭✭
    ✭✭
    Soooooo, since us poor guys on the east side of the big pond will still have to wait a few hours more (i.e. until tomorrow), how's it going for you on the other side? Today's patch notes covered pretty much anything I had on my list.
  • Delsskia
    Delsskia
    ✭✭✭✭
    I logged on and the first thing I saw was a pink texture... So that's not fixed. But now it's on plants rather than pets and mounts.
    NA-PC
    Fantasia
  • Randay
    Randay
    ✭✭✭
    You sure, that you didn't eat the wrong plants?:D
    Edited by Randay on March 14, 2016 4:44PM
    if less really is more, then maybe nothing is everything
  • tfraley
    tfraley
    Been able to play without crashing which is so nice.

    Still getting pink textures and occasional parts missing on the map
Sign In or Register to comment.