Jump to content

HitmanFF

Member
  • Content Count

    647
  • Joined

  • Last visited

  • Medals

  • Medals

Posts posted by HitmanFF


  1. 2. 34 seconds ... how much is it on the Win32 Dedi Server? Have you got some comparison?

    With squad xml it's 4-5 seconds on a win32 server, 34 seconds on a linux server (latest build).

    It seems that the server doesn't freeze up completely anymore, but the long waiting time is still there.


  2. @The problem with squad.xml:

    Sounds like a problem in resolving the hostnames of the squad xml's.

    [*] Are the nameservers that are being used by the linux server responding fast enough and not taking seconds or longer to respond?

    [*] Is the server actually able to resolve hostnames?

    [*] In any case, using the NSCD (altough I have no experience nor much info on it), or caching features within a local DNS server, might help incase of squad.xml's from different players, hosted on the same websites

    Unfortunately it doesn't seem to be the case of faulty or slow name resolving. The main server I'm using has a vmware + windows XP setup on top of linux, so both the Windows and linux arma servers are using identical DNS servers with alsmost identical connection parameters.

    The 2nd test box has a (caching) DNS server running, and it's behaving exactly the same.


  3. A member was already gaming and i was joining the session, when he got not even the chain symbol (red or yellow), but experienced the same effect, as if the ofpserver (comparable) has crashed: having all the playerobjects running on the same position without really moving.

    My connection process on the other hand went fine as usual. When i was in the game and connected, his session was also continuing.

    Has anyone experienced similar effects?

    Yes, I indeed experienced the same issue. Server performance dropped to 0 fps for a short while (during the connection phase of the player with squad xml), which would make it seem like the server had crashed.


  4. But the param -cfg= is meant to be used for a serverconfig and not the arma.cfg?!

    The server is creating an arma.cfg like it is creating the user directory, but it is somewhere down the street and not actually inside the ArmA directory. But this is maybe an outcome of the fact, that the admin is currently running as root smile_o.gif

    The one and only userconfigfile one can use is the whatever/player/player.armaprofile/ ??

    Edit: No seems not. We put an additional subdir into it, a user big_bad_admin at first but it seems not to work.

    The server then creates a directory called big_bad_admin in the ArmA root directory when it was started with the param -name= ...

    According the biki on start up parameters, there is both a -config= for the server config and a -cfg= for the arma.cfg.

    The arma.cfg created by arma is in the / (server root directory) of the server, probably indeed because I was testing while being root.


  5. The player/player.armaprofile is working fine on my setup; the difficulties set in there are effective.

    I'm having serious issues with (lack of) bandwidth usage though. According the biki, you should set those in arma.cfg.

    But one question remains: where is the third configfile, the arma.cfg?

    As I was unsure of its default location, I've added <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">-cfg=arma.cfg to the startup command line.

    But connecting to the server takes a very long time (from xxx connecting to xxx connected will take well over 50 seconds, whereas it's 5 seconds max on the Windows equivalent), and I'm getting multiple 'network message pending' messages, which I assume also are related to bandwidth:<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">Server: Network message c737 is pending

    Server: Network message c737 is pending

    Server: Network message c737 is pending

    Server: Network message c737 is pending

    Server: Network message c737 is pending

    Server: Network message c737 is pending

    Server: Network message c758 is pending

    Server: Network message c759 is pending

    Server: Network message c75a is pending

    Server: Network message c75b is pending

    .

    We've only tested the server for a short while last night, since after less than 30 minutes it turned inresponsive (all other players ran on the spot, unable to kill them), though there was no 'session lost', red link or any other sign of server disappearing.


  6. Followed the instructions, started the server and only got <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">Floating point exception

    Linux distro is Ubuntu 6.06 LTS (kernel 2.6.15-7) 32 bits, on an Intel E6600 dual core.


  7. There are two omissions in your drop script, as far as I can see. _weapon is not assigned any value, and you will always pass the #Droplaw part, because you don't jump over it in case the weapon is not the I44_PzFaust.

    I can't test it for you, but this should work better:

    <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">_unit =_this select 0

    _weapon = _this select 1

    ~1

    ?(_weapon == "I44_PzFaust"):goto "DropLaw"

    goto "exit"

    #DropLaw

    _unit action ["Drop Weapon",_unit,0,0,"I44_PzFaust"]

    goto "exit"

    #exit

    exit

    One thing to consider is to remove all I44_PzFaust ammo from the player (at least if you're able to carry more than one rocket for it) after firing. Otherwise you'd be able to pick up the Paunzerfaust from the ground, reload and fire it again.


  8. I tried the following script, and it worked in MP on a dedicated server:

    test.sqs:

    <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">@ (time > 1)

    hint format["%1", typeOf player]

    ~10

    _newunit = group player createunit["SquadLeaderE", position player, [], 0, "FORM"]

    selectPlayer _newunit

    hint format["%1", typeOf player]

    ~10

    _newunit = group player createunit["SoldierGNOG", position player, [], 0, "FORM"]

    selectPlayer _newunit

    hint format["%1", typeOf player]

    <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">[this] exec "test.sqs"in the init field of the initial unit.


  9. Before the isServer command was available, you could check by creating a game logic (in your question called server) on the map and check for its availability by <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">?!(local server)

    So if the logic is available, it means it's the server.

    The isServer command has made this construction redundant basically.


  10. The ofp server needs all file names (pbo's etc) to be in lower case. Try to compile tolower.c with <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">make tolower, then run <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">./tolowerfrom the ofp directory.

    Then try running the ofp server again.

    Good luck wink_o.gif


  11. There's one major draw back for me, and that's the fact that it has to be integrated in the map. That means for me that responsibility is both on the server admin (for making the server sided addon available on the server) and on the map maker (for adding the call). The map maker would either be making a map that cannot be run on servers that don't have the addon present, or he'd have to add the option to enable the frame work or not (e.g. through param1/param2 selection).

    If it was possible to run the frame work as part of the server sided scripting, e.g. on a onUserConnected event, I'd definitely say yes. That way the map maker is not burdened with integrating the framework (even if it's just one single command to add), and the server admin would not have to limit his choice of maps to those using the frame work.


  12. Start the server binary, not the script and he will tell you the reason he aborts.

    Please try and start the server binary directly, to see what's happening:<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">cd /home/gr/ofp

    ./server -config=server.cfg -port=2302and check for any output.

    The start of this thread has a number of solutions for the ofp server not starting up as it should. glibc version, server's host name and dns resolution issues are the most common causes for servers crashing/dying.

×