Tag - GitHub

1~2/2

  • Docker로 Palworld 서버 열기 (ARM64) Ver. 2

    2024-02-09 21:34:54 Palworld 서버를 24시간 열어둘 경우 생기는 문제 어제 위 글을 작성하고 또 무슨 기능이 생겼나 구경하려고 들어갔는데.. ARM 버전이 생겼다? 그것도 FEX-Emu를 사용하지 않는다?? 이게 되면 바로 갈아탈 마음에 바로 사용해 봤다. 실행 docker-compose.yml 파일 작성 후 docker compose up -d 실행 version: "3.9" services: palworld: container_name: palworld restart: unless-stopped stop_grace_period: 30s image: thijsvanloef/palworld-server-docker:latest-arm64 ports: - 8211:8211/udp - 27015:27015/udp env_file: - .env - settings.env environment: - PUID=1001 - PGID=1001 - MULTITHREADING=true - UPDATE_ON_BOOT=true - TZ=Asia/Seoul volumes: - ./server:/palworld networks: games: ipv4_address: 172.16.11.11 networks: games: name: games external: true 이 글에서 image 값 끝에 -arm64 추가한 것 외엔 차이가 없다. PUID, PGID는 클라우드로 옮기면서 변경한 것. 문제 발생 처음엔 잘 되는가 싶더니 컨테이너를 그냥 재시작만 했는데 에러가 났다. docker logs -f palworld로 확인한 결과 재시작을 할 때마다 정상적으로 돌아갈 때도 있고, 아래처럼 로그가 찍히면서 혼자 재시작을 하기도 한다. ... *****STARTING SERVER***** ./PalServer-arm64.sh -port=8211 -queryport=27015 EpicApp=PalServer -useperfthreads -NoAsyncLoadingThread -UseMultithreadForDS time="2024-02-08T20:20:05+09:00" level=info msg="read crontab: /home/steam/server/crontab" [S_API] SteamAPI_Init(): Loaded local 'steamclient.so' OK. Shutdown handler: initalize. Increasing per-process limit of core file size to infinity. - Existing per-process limit (soft=18446744073709551615, hard=18446744073709551615) is enough for us (need only 18446744073709551615) CAppInfoCacheReadFromDiskThread took 24 milliseconds to initialize Steam Service Error: Failed to get Steam Service Start function Setting breakpad minidump AppID = 2394010 [S_API FAIL] Tried to access Steam interface SteamUser021 before SteamAPI_Init succeeded. [S_API FAIL] Tried to access Steam interface SteamFriends017 before SteamAPI_Init succeeded. [S_API FAIL] Tried to access Steam interface STEAMAPPS_INTERFACE_VERSION008 before SteamAPI_Init succeeded. [S_API FAIL] Tried to access Steam interface SteamNetworkingUtils004 before SteamAPI_Init succeeded. free(): invalid next size (normal) Signal 6 caught. Malloc Size=262146 LargeMemoryPoolOffset=262162 Malloc Size=131160 LargeMemoryPoolOffset=393352 Malloc Size=131160 LargeMemoryPoolOffset=524536 LogPakFile: Display: Found Pak file ../../../Engine/Programs/CrashReportClient/Content/Paks/CrashReportClient.pak attempting to mount. LogPakFile: Display: Mounting pak file ../../../Engine/Programs/CrashReportClient/Content/Paks/CrashReportClient.pak. LogPakFile: Display: Mounted Pak file '../../../Engine/Programs/CrashReportClient/Content/Paks/CrashReportClient.pak', mount point: '../../../Engine/' LogICUInternationalization: ICU TimeZone Detection - Raw Offset: +9:00, Platform Override: '' LogInit: Build: ++UE5+Release-5.1-CL-0 LogInit: Engine Version: 5.1.1-0+++UE5+Release-5.1 LogInit: Compatible Engine Version: 5.1.0-0+++UE5+Release-5.1 LogInit: Net CL: 0 LogInit: OS: Debian GNU/Linux 11 (bullseye) (5.15.0-1052-realtime), CPU: Box64 on Neoverse-N1 @1000 MHz, GPU: UnknownVendor PCI-id: 108e-0010 LogInit: Compiled (64-bit): Dec 31 2023 20:12:22 LogInit: Compiled with Clang: 13.0.1 (https://github.com/llvm/llvm-project 75e33f71c2dae584b13a7d1186ae0a038ba98838) LogInit: Build Configuration: Shipping LogInit: Branch Name: ++UE5+Release-5.1 LogInit: Command Line: -Abslog="/palworld/Pal/Saved/Logs/Pal-CRC.log" -Unattended -ImplicitSend "/palworld/Pal/Saved/Crashes/crashinfo-Pal-pid-111-AE37D7ED12904D579713C0111B657EBB/" -unattended LogInit: Base Directory: /palworld/Engine/Binaries/Linux/ LogInit: Allocator: Mimalloc LogInit: Installed Engine Build: 1 LogInit: Presizing for max 100000 objects, including 0 objects not considered by GC, pre-allocating 0 bytes for permanent pool. LogInit: Object subsystem initialized LogConfig: Applying CVar settings from Section [ConsoleVariables] File [Engine] [2024.02.08-11.20.29:881][ 0]LogInit: Unix hardware info: [2024.02.08-11.20.29:881][ 0]LogInit: - we are the first instance of this executable [2024.02.08-11.20.29:881][ 0]LogInit: - this process' id (pid) is 167, parent process' id (ppid) is 111 [2024.02.08-11.20.29:882][ 0]LogInit: - we are not running under debugger [2024.02.08-11.20.29:882][ 0]LogInit: - machine network name is 'a4f09a0272ae' [2024.02.08-11.20.29:882][ 0]LogInit: - user name is 'steam' (steam) [2024.02.08-11.20.29:883][ 0]LogInit: - we're logged in locally [2024.02.08-11.20.29:883][ 0]LogInit: - we're running with rendering [2024.02.08-11.20.29:883][ 0]LogInit: - CPU: GenuineIntel 'Box64 on Neoverse-N1 @1000 MHz' (signature: 0x601) [2024.02.08-11.20.29:883][ 0]LogInit: - Number of physical cores available for the process: 4 [2024.02.08-11.20.29:883][ 0]LogInit: - Number of logical cores available for the process: 4 [2024.02.08-11.20.29:883][ 0]LogInit: - Cache line size: 64 [2024.02.08-11.20.29:883][ 0]LogInit: - GPU Brand Info: UnknownVendor PCI-id: 108e-0010 [2024.02.08-11.20.29:883][ 0]LogInit: - Memory allocator used: Mimalloc [2024.02.08-11.20.29:884][ 0]LogInit: - This binary is optimized with LTO: no, PGO: no, instrumented for PGO data collection: no [2024.02.08-11.20.29:884][ 0]LogInit: - This is an internal build. [2024.02.08-11.20.29:885][ 0]LogCore: Skipped benchmarking clocks because the engine is running in a standalone program mode - CLOCK_MONOTONIC will be used. [2024.02.08-11.20.29:885][ 0]LogInit: Unix-specific commandline switches: [2024.02.08-11.20.29:885][ 0]LogInit: -ansimalloc - use malloc()/free() from libc (useful for tools like valgrind and electric fence) [2024.02.08-11.20.29:885][ 0]LogInit: -jemalloc - use jemalloc for all memory allocation [2024.02.08-11.20.29:885][ 0]LogInit: -binnedmalloc - use binned malloc for all memory allocation [2024.02.08-11.20.29:885][ 0]LogInit: -filemapcachesize=NUMBER - set the size for case-sensitive file mapping cache [2024.02.08-11.20.29:885][ 0]LogInit: -useksm - uses kernel same-page mapping (KSM) for mapped memory (OFF) [2024.02.08-11.20.29:885][ 0]LogInit: -ksmmergeall - marks all mmap'd memory pages suitable for KSM (OFF) [2024.02.08-11.20.29:885][ 0]LogInit: -preloadmodulesymbols - Loads the main module symbols file into memory (OFF) [2024.02.08-11.20.29:885][ 0]LogInit: -sigdfl=SIGNAL - Allows a specific signal to be set to its default handler rather then ignoring the signal [2024.02.08-11.20.29:885][ 0]LogInit: -crashhandlerstacksize - Allows setting crash handler stack sizes (204800) [2024.02.08-11.20.29:885][ 0]LogInit: -noexclusivelockonwrite - disables marking files created by the engine as exclusive locked while the engine has them opened [2024.02.08-11.20.29:885][ 0]LogInit: -httpproxy=ADDRESS:PORT - redirects HTTP requests to a proxy (only supported if compiled with libcurl) [2024.02.08-11.20.29:886][ 0]LogInit: -reuseconn - allow libcurl to reuse HTTP connections (only matters if compiled with libcurl) [2024.02.08-11.20.29:886][ 0]LogInit: -virtmemkb=NUMBER - sets process virtual memory (address space) limit (overrides VirtualMemoryLimitInKB value from .ini) [2024.02.08-11.20.29:886][ 0]LogInit: - Physical RAM available (not considering process quota): 24 GB (23989 MB, 24564872 KB, 25154428928 bytes) [2024.02.08-11.20.29:886][ 0]LogInit: - VirtualMemoryAllocator pools will grow at scale 1.4 [2024.02.08-11.20.29:886][ 0]LogInit: - MemoryRangeDecommit() will be a no-op (re-run with -vmapoolevict to change) [2024.02.08-11.20.29:886][ 0]LogInit: - PageSize 4096 [2024.02.08-11.20.29:886][ 0]LogInit: - BinnedPageSize 65536 [2024.02.08-11.20.30:006][ 0]LogUObjectArray: 419 objects as part of root set at end of initial load. [2024.02.08-11.20.30:006][ 0]LogUObjectAllocator: 89056 out of 0 bytes used by permanent object pool. [2024.02.08-11.20.30:006][ 0]LogUObjectArray: CloseDisregardForGC: 0/0 objects in disregard for GC pool [2024.02.08-11.20.30:012][ 0]LogPaths: Warning: No paths for game localization data were specifed in the game configuration. [2024.02.08-11.20.30:013][ 0]LogInit: Using OS detected language (en-US-POSIX). [2024.02.08-11.20.30:013][ 0]LogInit: Using OS detected locale (en-US-POSIX). [2024.02.08-11.20.30:014][ 0]LogInit: Warning: No paths for engine localization data were specifed in the engine configuration. [2024.02.08-11.20.30:020][ 0]LogTextLocalizationManager: No localization for 'en-US-POSIX' exists, so 'en' will be used for the language. [2024.02.08-11.20.30:021][ 0]LogTextLocalizationManager: No localization for 'en-US-POSIX' exists, so 'en' will be used for the locale. [2024.02.08-11.20.30:026][ 0]LogInit: Using OS detected language (en-US-POSIX). [2024.02.08-11.20.30:026][ 0]LogInit: Using OS detected locale (en-US-POSIX). [2024.02.08-11.20.30:026][ 0]LogTextLocalizationManager: No localization for 'en-US-POSIX' exists, so 'en' will be used for the language. [2024.02.08-11.20.30:026][ 0]LogTextLocalizationManager: No localization for 'en-US-POSIX' exists, so 'en' will be used for the locale. [2024.02.08-11.20.30:043][ 0]LogPackageLocalizationCache: Processed 2 localized package path(s) for 1 prioritized culture(s) in 0.004851 seconds [2024.02.08-11.20.30:046][ 0]CrashReportCoreLog: CrashReportClientVersion=1.0 [2024.02.08-11.20.30:046][ 0]CrashReportCoreLog: CrashReportReceiver disabled [2024.02.08-11.20.30:047][ 0]CrashReportCoreLog: DataRouterUrl: https://o1291919.ingest.sentry.io/api/6513339/unreal/4a1a3921f51f4975b4cf8dd19022cb20/ [2024.02.08-11.20.30:080][ 0]CrashReportCoreLog: Initial state = Unknown UploadState value [2024.02.08-11.20.30:080][ 0]CrashReportCoreLog: Initial state = Unknown UploadState value [2024.02.08-11.20.30:096][ 0]LogCrashDebugHelper: DepotName: //UE5/Release-5.1 [2024.02.08-11.20.30:096][ 0]LogCrashDebugHelper: BuiltFromCL: 0 [2024.02.08-11.20.30:096][ 0]LogCrashDebugHelper: EngineVersion: 5.1.1-0+++UE5+Release-5.1 [2024.02.08-11.20.30:096][ 0]LogCrashDebugHelper: BuildVersion: ++UE5+Release-5.1-CL-0 [2024.02.08-11.20.31:081][ 0]CrashReportCoreLog: Got 3 pending files to upload from 'crashinfo-Pal-pid-111-AE37D7ED12904D579713C0111B657EBB' [2024.02.08-11.20.31:081][ 0]CrashReportCoreLog: State change from Ready to SendingFiles [2024.02.08-11.20.31:081][ 0]CrashReportCoreLog: CompressAndSendData have 3 pending files [2024.02.08-11.20.31:082][ 0]CrashReportCoreLog: CompressAndSendData compressing 152 bytes ('/palworld/Pal/Saved/Crashes/crashinfo-Pal-pid-111-AE37D7ED12904D579713C0111B657EBB/CrashReportClient.ini') [2024.02.08-11.20.31:083][ 0]CrashReportCoreLog: CompressAndSendData compressing 4518 bytes ('/palworld/Pal/Saved/Crashes/crashinfo-Pal-pid-111-AE37D7ED12904D579713C0111B657EBB/CrashContext.runtime-xml') [2024.02.08-11.20.31:083][ 0]CrashReportCoreLog: CompressAndSendData compressing 385 bytes ('/palworld/Pal/Saved/Crashes/crashinfo-Pal-pid-111-AE37D7ED12904D579713C0111B657EBB/Diagnostics.txt') [2024.02.08-11.20.31:206][ 0]LogInit: Using libcurl 7.83.1 [2024.02.08-11.20.31:206][ 0]LogInit: - built for Linux [2024.02.08-11.20.31:207][ 0]LogInit: - supports SSL with OpenSSL/1.1.1n [2024.02.08-11.20.31:207][ 0]LogInit: - supports HTTP deflate (compression) using libz 1.2.12 [2024.02.08-11.20.31:207][ 0]LogInit: - other features: [2024.02.08-11.20.31:207][ 0]LogInit: CURL_VERSION_SSL [2024.02.08-11.20.31:207][ 0]LogInit: CURL_VERSION_LIBZ [2024.02.08-11.20.31:207][ 0]LogInit: CURL_VERSION_IPV6 [2024.02.08-11.20.31:207][ 0]LogInit: CURL_VERSION_ASYNCHDNS [2024.02.08-11.20.31:208][ 0]LogInit: CURL_VERSION_LARGEFILE [2024.02.08-11.20.31:212][ 0]LogInit: CurlRequestOptions (configurable via config and command line): [2024.02.08-11.20.31:212][ 0]LogInit: - bVerifyPeer = false - Libcurl will NOT verify peer certificate [2024.02.08-11.20.31:212][ 0]LogInit: - bUseHttpProxy = false - Libcurl will NOT use HTTP proxy [2024.02.08-11.20.31:213][ 0]LogInit: - bDontReuseConnections = false - Libcurl will reuse connections [2024.02.08-11.20.31:213][ 0]LogInit: - MaxHostConnections = 16 - Libcurl will limit the number of connections to a host [2024.02.08-11.20.31:213][ 0]LogInit: - LocalHostAddr = Default [2024.02.08-11.20.31:213][ 0]LogInit: - BufferSize = 65536 [2024.02.08-11.20.31:221][ 0]CrashReportCoreLog: Sending HTTP request: https://o1291919.ingest.sentry.io/api/6513339/unreal/4a1a3921f51f4975b4cf8dd19022cb20/?AppID=CrashReporter&AppVersion=5.1.1-0%2B%2B%2BUE5%2BRelease-5.1&AppEnvironment=Release&UploadType=crashreports&UserID=-000003e9%7C%7C [2024.02.08-11.20.31:563][ 0]CrashReportCoreLog: OnProcessRequestComplete(), State=SendingFiles bSucceeded=1 [2024.02.08-11.20.31:563][ 0]CrashReportCoreLog: State change from SendingFiles to SendingFiles [2024.02.08-11.20.31:564][ 0]CrashReportCoreLog: All uploads done [2024.02.08-11.20.31:564][ 0]CrashReportCoreLog: State change from SendingFiles to Finished [2024.02.08-11.20.32:130][ 0]CrashReportCoreLog: Final state (Receiver) = Finished [2024.02.08-11.20.32:130][ 0]CrashReportCoreLog: Final state (Receiver) = Unknown UploadState value [2024.02.08-11.20.32:155][ 0]LogCore: Engine exit requested (reason: CrashReportClientApp RequestExit) [2024.02.08-11.20.32:155][ 0]LogExit: Preparing to exit. [2024.02.08-11.20.32:192][ 0]LogExit: Object subsystem successfully closed. [2024.02.08-11.20.32:196][ 0]LogModuleManager: Shutting down and abandoning module HTTP (12) [2024.02.08-11.20.32:233][ 0]LogModuleManager: Shutting down and abandoning module SSL (11) [2024.02.08-11.20.32:233][ 0]LogModuleManager: Shutting down and abandoning module CrashDebugHelper (8) [2024.02.08-11.20.32:237][ 0]LogModuleManager: Shutting down and abandoning module CoreUObject (6) [2024.02.08-11.20.32:237][ 0]LogModuleManager: Shutting down and abandoning module PakFile (4) [2024.02.08-11.20.32:238][ 0]LogModuleManager: Shutting down and abandoning module RSA (3) [2024.02.08-11.20.32:247][ 0]LogExit: Exiting. CommonUnixCrashHandler: Signal=6 Engine crash handling finished; re-raising signal 6 for the default handler. Good bye. 또는 18번 라인까지만 뜨면서 재시작이 안될 때도 있는데, 이건 그냥 멈춰버린 듯. 접속 시 타임아웃이 발생한다. 해결 (처음 올린 GitHub Issue) 그냥 원래 쓰던 빌드한 이미지로 쓸까 하다가 그냥 Issues에 한 번 올려나 봤다. # 그리고 답글이 달렸고, 해결됐다! 해결 방법은 environment: - BOX64_DYNAREC_STRONGMEM=3 - BOX64_DYNAREC_BIGBLOCK=0 - BOX64_DYNAREC_BLEEDING_EDGE=0 이런 환경변수를 추가하는 것으로, 최종 docker-compose.yml 파일은 아래와 같다. version: "3.9" services: palworld: container_name: palworld restart: unless-stopped stop_grace_period: 30s image: thijsvanloef/palworld-server-docker:latest-arm64 ports: - 8211:8211/udp - 27015:27015/udp env_file: - .env - settings.env environment: - PUID=1001 - PGID=1001 - MULTITHREADING=true - UPDATE_ON_BOOT=true - TZ=Asia/Seoul - BOX64_DYNAREC_STRONGMEM=3 - BOX64_DYNAREC_BIGBLOCK=0 - BOX64_DYNAREC_BLEEDING_EDGE=0 volumes: - ./server:/palworld networks: games: ipv4_address: 172.16.11.11 networks: games: name: games external: true FEX-Emu를 사용하지 않기 때문에 이 문제도 없으니 이제 나도 이 이미지의 모든 기능들을 귀찮게 새로 빌드하지 않고 쉽게 사용할 수 있게 됐다..!
    Game > Palworld

  • Palworld 서버를 24시간 열어둘 경우 생기는 문제

    2024-02-08 18:04:01 어제 도커로 서버 올려놓고 닫지 않고 잤더니 애들이 다 죽기 직전이네.. 원인 플레이어가 팰 근처인지 거점 근처인지 거기서부터 멀어질수록 얘들이 하는 행동이 느려진다고 한다. 그래서 서버에 아무도 없으면 전혀 움직이지 않아서 생기는 문제라고.. 임시 대처 방법 도커 이미지 GitHub의 Issues에 해당 문제가 올라와 있다. 해당 페이지의 하단에 어느 한국인이 파이썬 스크립트를 짜놓은 게 있는데, rcon을 통해 접속 중인 플레이어가 있는지 체크하고, 없다면 컨테이너 자체를 중지한 후 tcpdump 명령어로 서버 연결 포트를 탐지하여 연결 요청이 들어오면 다시 컨테이너가 돌아가게 만드는 코드다. 코드는 컨테이너가 아닌 호스트에서 실행해야 한다. 아래 코드는 조금 수정한 코드. import subprocess import time TARGET_PORT = 8211 CONTAINER_NAME = "palworld-server" sudo_password = "your_sudo_password_here" # sudo 명령어 사용 시 패스워드 입력이 필요없다면 삭제 def check_players(): players_output = run_command( f"docker exec {CONTAINER_NAME} rcon-cli ShowPlayers" ) print(f"player: {players_output}") return "," in players_output.replace("name,playeruid,steamid", "") # itzg/rcon-cli을 사용할 때 문제 발생 해결 def run_command(command): """Helper function to run a shell command and return its output""" result = subprocess.run( command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True, text=True ) return result.stdout.strip() def main(): if not check_players(): print("No players found. Restart the server.") run_command(f"docker restart {CONTAINER_NAME}") for sec in range(30, 0, -1): print(f"Waiting for restart to complete... {sec} seconds left") time.sleep(1) # if not players, pause server if not check_players(): print("Restart complete. Pause the server.") run_command(f"docker pause {CONTAINER_NAME}") player = False print("Waiting for players to connect") else: player = True while not player: # Use tcpdump to capture incoming traffic on the target port #capture_command = f"echo {sudo_password} | sudo -S tcpdump -n -c 1 -i any port {TARGET_PORT} 2>/dev/null" capture_command = f"sudo -S tcpdump -n -c 1 -i any port {TARGET_PORT} 2>/dev/null" # 패스워드 입력 필요없는 경우 capture_result = run_command(capture_command) # Check if any packets were captured if capture_result: player = True print(f"Connection attempt detected on port {TARGET_PORT}") print("Starting server") run_command(f"docker unpause {CONTAINER_NAME}") for sec in range(30, 0, -1): print(f"Waiting for players to connect... {sec} seconds left") time.sleep(1) else: time.sleep(1) else: print("Players found. Server will continue running.") time.sleep(60) if __name__ == "__main__": while True: main() 코드를 수정한 이유는 이전 글에서 말한 itzg/rcon-cli을 사용할 때 뜨는 문구 때문에 check_players 함수는 항상 True를 리턴함. tcpdump 명령어가 요청을 대기하지 않고 그냥 넘어가버림. 이 글을 쓰는 지금은 1번의 경우 rcon을 변경해서 상관없지만, 2번은 sudo 명령어 사용 시 패스워드 입력이 필요 없어서 발생한 문제다. 패스워드 입력을 위해 echo {sudo_password} | 명령어를 사용하는데, 패스워드를 묻지 않기 때문에 표준 입력으로 넘긴 저 패스워드가 tcpdump 명령어가 받아서 그런 듯. 수정한 부분은 해당 issue에 댓글로 남겨놨다 사용법 위 코드를 파일에 저장하고 python3 file.py 명령어로 스크립트 실행. 또는 터미널을 닫아도 백그라운드로 돌아가게 하려면 nohub python3 file.py 2>&1 & 명령어 실행. 잘만 하면 컨테이너 내부에서 실행하도록 만들어서 이미지 빌드 할 때 기능으로 추가해도 될 듯한데..
    Game > Palworld