I've found the instance of my server session wasn't running the mod for some reason. I mustn't have saved the start up powershell script when adding the command parameter before executing it or something. However, the problem still stands... this is an absolutely mind-boggling oversight that Bohemia can release an update without the basic respawn functionality working, and don't release the patch to rectify the issue when they identify it.

Another strange thing is that when in debug mode, if I look at the office building and use "CursorTarget", it returns "ObjNull", yet when using "CursorObject", the command returns the offices.
Using CursorTarget above in specifying the bis_disabled_door variable, it still assigned it...

It seems that the "nearestBuilding" command doesn't recognise the Land_Office building anymore. Why?

I've done some further testing on this.
When using "CursorTarget" instead of the buildingName, the disabled door parameter works.

In case anyone finds this thread looking for the same answers...

@56curious - IF you took the time to read my post, you'd see that I still want VON enabled.
I already have a makeshift script to disable it, but as soon as any client lag is experienced, the makeshift "fixes" do not work... hence why I have made this ticket to actually have an inbuilt option to remove the voice from created channels using the create custom radio channel commands.

Unless str is inside the compile command, then it isn't used.

Is that the same for how diag_log shows that information?
Because all of my database queries now have ""2"" etc around numbers from selected lists where as before in diag_log they displayed as 2...
For those fields, I don't have any of them using str() they're loaded and used as integers once compiled from the string format in the query.

That one element might be fixed, but now if you compile "2" it becomes {""2""} instead of {2}... why?

Nothing currently works in my Altis Life server when reading from the database anymore...
It shouldn't have been changed. If it ain't broke, don't ""fix"" it.

This has broken basic functionality of the compile command. String arrays "["text","text,"text"]" now compile as [""text"",""text"",""text""] and throw errors because it is not written as "",text,"","",text, etc...

This was due to having over 4,000 custom placed objects across the map of Altis...
It would be beneficial to have some kind of memory limiter built in to the editor to say "You've gone over the usable limit of place-able objects."

I bypassed this by retrieving the player's UID first, and then running at the query of all follow up scripts against all player's UIDs and then matching it back to the correct player.

On the new server, with better memory, storage and processing, the server performs even worse.
It takes about 10 minutes for it to even appear in the server browser, which suggests the problem is with the arma3server release for 1.60.

The server runs on x86 as well, which is bad. He process itself usually runs quite high in its maximum utilisation of 2047MB (as malloc describes anyway). Not to mention the recent build is full of memory crashes for dedicated servers...
X64 processing is always going to be better and faster.

It's just not going to be feasible with Arma3.

timeoutDisconnect has been removed and the issue is still persisting...
Various clients are connected through various performance development executables - basic/v16/v17/v19

Server is running on v17, will be upgrading to v19 tonight.

New log attached.
Client using v19
Server using v19

Towards the end, the client receives 3 disconnects in about 15 minutes.
Server has been up for about 3 hours at this point...

This is potentially caused by texture files (not entirely sure how or why).
We stumbled across something about corrupt texture files and how they're causing the server to crash, and we'd just updated a texture before the issue started occurring to everyone (note, the issue was apparent before the update of textures).
I've since removed the texture and excluded it from the scripts and we haven't had anyone kicked to side select yet.

Still getting the issue.
Will remove the timeoutDisconnect option on the next reboot.

The issue randomly fixed itself... After booting 15 people back to side select screen every 5-10 minutes, after the server was online for 90 minutes it just stopped kicking everyone and we changed nothing.

This leads me to believe it's potentially a networking issue or a performance issue. It still could have something to do with the new config parameter... Should the config parameter show a "disconnected" message when if removes an inactive client?

For the record, we're now using v17 perf files and timeoutDisconnect is set to 120.

I have the following set in Config_server.cfg;
disconnectTimeout = 30;

I noticed on Dwarden's post in the server perf binaries that this is a new command. Would it be this that could be causing the disconnect to side-select without stating the player has disconnected?

Added two different client RPT files and the server RPT files.
There's approximately 12 different people experiencing the same issues here now...

Something that is concerning is usually when a player disconnects, it displays "player has disconnected" in systemChat.
Through this problem, it is not logging a player disconnect in systemChat either...

I have no scripts that I know of that would remove someone to the side select screen (or abort them, which is essentially what's happening...).

Does indexing start at 0 or 1?

ok, cheers.

When the channels are first created, they're created empty, and players are adding into this on joining the server.

On server;
life_radio_west = radioChannelCreate [[0, 0, 0.8, 0.8], "Police Channel", "%UNIT_NAME", [], false]; ID 6
life_radio_civ = radioChannelCreate [[0.5, 0, 0.6, 0.8], "Civilian Channel", "%UNIT_NAME", [], false]; ID 7
life_radio_indep = radioChannelCreate [[0, 1, 0.2, 0.8], "EMS Channel", "%UNIT_NAME", [], false]; ID 8
life_radio_rebel = radioChannelCreate [[0.5,0,0,0.8], "Rebel Channel", "%UNIT_NAME", [], false]; ID 9
life_radio_corp = radioChannelCreate [[0,0.8,0,0.8], "Army Channel", "%UNIT_NAME", [], false]; ID 10
life_radio_bh = radioChannelCreate [[0, 0.95, 1, 0.8], "Bounty Hunter Channel", "%UNIT_NAME", [], false]; ID 11
life_radio_global = radioChannelCreate [[0.8,0.8,0.8,0.8], "Global Channel", "%UNIT_NAME", [], false]; // ID 12

On server when player joins (inside switch based on playerside);
case west:
{

		if(_bool) then
		{
			life_radio_west radioChannelAdd [_unit];
			life_radio_global radioChannelAdd [_unit];
		}
			else
		{
			life_radio_west radioChannelRemove [_unit];
			life_radio_global radioChannelRemove [_unit];
		};

};

The same section of code also handles the removal, should the correct circumstances occur.

I've tried several ways; using all custom channels I've created 6-12, different test strings, tried sending text through too, used player as the object, a gamelogic, a player assigned in a local variable, global variable... the actual player name.

None have returned any text in the channels.
I'm basically setting up a scenario where when a player accepts a task assignment, and they automatically post an update to their commander where they're heading for the task.

Yep, channel 6 effectively replaces the side channel for West, but also adds in a couple of other players and renames it.
Players can talk freely in this channel (and all other custom channels, upto 12).
Each ID was obtained from when they were created on the server, but I've also confirmed the channel numbers by selecting the custom radio channel and then opening the debug console and typing hint format["%1",currentChannel];

....really need this fixing as it greatly breaks Altis Life servers having direct communication channel unusable at a distance of 10m.....

Hi, sorry, we already have.
We've used perf v7, v8 and v9

Added an RPT log from the new rebuild... first time having this type of error, no idea what the issue is.

May be moving to a higher spec server soon to see if this combats the issue.

To just add a little more feedback... these crashes only seem to happen from 6pm GMT through to about 11pm GMT. We've just had a crash there where only 4 players were online, yet earlier in the day with more than 4 players online we completed multiple 4 hour cycles without incident.

Reinstalled Arma3server on dedicated server earlier this morning from a complete fresh build.
It seems ok so far with <10 players on, it's made a 4 hour session without failing but still seeing various errors in RPT logs.

16:59:55 Server: Object 15:2 not found (message Type_119)
16:59:55 Server: Object 15:3 not found (message Type_119)
16:59:55 Server: Object 20:24 not found (message Type_119)
16:59:57 Server: Object 20:2 not found (message Type_119)
16:59:57 Server: Object 16:2 not found (message Type_119)
16:59:57 Server: Object 19:3 not found (message Type_119)
16:59:57 Server: Object 21:2 not found (message Type_119)
16:59:57 Server: Object 19:3 not found (message Type_91)
16:59:57 Server: Object 19:2 not found (message Type_91)
16:59:57 Server: Object 19:2 not found (message Type_119)
16:59:57 Server: Object 21:3 not found (message Type_119)
16:59:58 Server: Object 21:3 not found (message Type_91)
16:59:58 Server: Object 21:2 not found (message Type_91)
16:59:59 Server: Object 20:4 not found (message Type_119)
16:59:59 Server: Object 10:45 not found (message Type_119)
17:00:00 Server: Object 16:32 not found (message Type_91)

13:06:04 NetServer::SendMsg: cannot find channel #96071318, users.card=3
13:06:04 NetServer: users.get failed when sending to 96071318
13:06:04 Message not sent - error 0, message ID = ffffffff, to 96071318 (BeastOfEvolution)

13:45:57 Server: Network message 1417cf is pending
13:45:57 Server: Network message 1417d0 is pending
13:45:57 Server: Network message 1417d1 is pending
13:46:00 Server: Network message 141bfc is pending
13:46:00 Server: Network message 141bfc is pending
13:46:00 Server: Network message 141bfc is pending
13:46:00 Server: Network message 141bfc is pending

The first and the third error messages we were seeing previously to today as well in the lead up to the server crash... the second error message repeats an awful lot at server startup when the player first tried connecting, but then didn't appear again after that.

server_crash.zip added...

Two crashes within 20 minutes of one another.

To clarify, the issue occurs on the server NOT the client side.
The server has the perfv7 binaries on it with the malloc provided in the same dropbox location as we constantly have issues with the server crashing since the latest arma updates...

It's any animation that's in a loop. It performs the first animation successfully and then the rest jitter between the first few frames (depending on the sleep duration inside the loop).

Considering you're actually responding to topics, can someone post an update on this: http://feedback.arma3.com/view.php?id=26878
There's literally hundreds of people waiting on some form of official communication from BI on this...

Hi, sorry... what do you mean by "repro" mission? The mission.pbo file I have won't work without the mods - @life_server and @extDB.

Got the same issue happening on windows 2012 dedicated server. 1.54, latest malloc file released by dev, 10-20 people online get between 10 minutes and 2 hours playtime at best.

Only started happening in last 2 days, been working fine since 23rd December before then.

bump...
any update from BI managers on where this is going?

Using v11... entire animation system completely crashed.
Could no longer perform any kind of restrain action in Altis Life.

v11 crash added.

Use the v9 perf if you're having problems with v8

Hmmm strange. Our server was online for a total of 12.5 hours yesterday before I took it down to test a mission file patch. There were still some object not found (message type_119/91) errors and after about 6 hours we got network message is pending...

Which Malloc files are you using with perf v8? I'm still on v7 perf malloc.
We haven't had any client crashes, but when we used perf v7 with v7 malloc we got crashes with less than 3 players after 10 minutes.

perf v8 seems to fix the issue. Had the server running (only myself in it) for 4 hours, no errors listed at all.

Still keeping our server password on though for now while some systems still get tweaked but you guys should download it and see how it fares with 30+ players.

https://www.dropbox.com/sh/582opsto4mmr8d8/3BSy9PdRGm

-ExtDB (35)
-life_server

Yeah, where's the support here?
Not that I'm the only one with this issue (obviously) but I've just spent the last 7 months building a server I can't use since that update. Roll the update back to a stable one and test fixing this issue in your own model office environment. Not on our clients and servers...

Where are the change logs?

...mine isn't, and I'm also using the v7 performance binary.

If the v8s fix the issue, how come they aren't being rolled out as standard?

Same issues on our server.
Relatively new server, only gets around 10-30 players on each cycle but still get the issue in some cases within 20 minutes.

Also noticed since this latest update, mission file is repeatedly "locked" when entering server after a mission update.

Just downloaded the latest performance version of arma3server.exe and the malloc.

I have the same issue in my mission. The eventHandler doesn't appear to fire when one player (as a medic or not) attempts to heal another player when using the FAK, but it does fire when healing self.

Yes, this is extremely annoying...

Just had my ticket closed for this ticket, and it's something that's 2+ years old.

Direct communication worked previously before 1.54... Now it's completely unusable.

I would love something like this included.

Also, it wouldn't have to be quite so complex as a gear system, but BI could just implement a setMaxVelocity command that allows us to define our own gear system, cruise control systems, or refine vehicle max speeds.

KeyHandlers can be used with that command to define manual gears, and other scripts could use the command to better restrict the vehicle for other uses.

The one thing I don't understand is, vehicles cap their own speed when driving on different terrain (hatchback for example drives only 27km/h off any type of road) but it also factors in hills... allowing scripters access to this would be amazing.

Is there a way to remove native wildlife? They serve no purpose and aren't synced... They should just be deleted.