terox

Patrol Op[s for Arma 3 is being released this weekend. We will be running an organised and managed session on both Saturday and Sunday at 18:30 hrs GMT (See first post for more details) If you dont know what Patrols Ops is, read more info here.... http://forums.bistudio.com/showthread.php?168315-A3-Patrol-Operations-Official-Thread This is a great public style mission played in a public environment in an organised and managed way. All sensible players who want to immerse into whatever role they select are more than welcome. If you are sick of wasteland and domination, come try it out

Zeus will be hosting Patrol Ops sessions this weekend Saturday & Sunday evenings, kickoff at 18:30 hrs GMT (You may also find us running sessions during the afternoons) So if you aren't in the bars getting legless pop by, all are welcome (Expect an organised and managed session, we work with an overall mission commander who gives orders to the squad leaders, who then carries them out with the help of the squad members and other support assets) More info here....... http://forums.bistudio.com/showthread.php?151063-Zeus-gaming-Nights

Am I the only one interested in the answer to this ?

For those of you who have received the dedi server package for testing. You may need to do the following to get your server to display in the server browser 1) Add the arma3server.exe to your firewall rules (its a different version of the.exe to the existing stable.exe 2) Copy over your battleye folder from your other server installs 3) edit your steamapp_id.txt from 233780 to 107410 4) Define a unique set of ports (Steam and Arma) and you should be good to auto install, you could run the following .cmd file @echo off SETLOCAL ENABLEDELAYEDEXPANSION ::STABLE VERSION SET A3BRANCH= 233780 SET A3Path=D:\Games\A3\A3Servertest SET STEAMLOGIN=loginname loginpassword echo TASK - Updating the Arma3 Dedicated server package install D:\Apps\steam\steamcmd.exe +login %STEAMLOGIN% +force_install_dir %A3Path% "+app_update %A3BRANCH%" validate +quit echo COMPLETED - A3 dedicated server package install updated echo .

There is an ongoing discussion about certain MP issues that we are having with server performance See this thread http://forums.bistudio.com/showthread.php?168337-Multiplayer-not-in-playable-state I have finished working on a benchmark mission to help reproduce a problem, where the server fps degrades to unplayable levels, in an environment where a fair sized number of players are present and a moderate number of AI exist at the start of the mission The problem I have, is it needs at least 30 players, ideally 40 or more to reproduce this I need your help if possible please This is a 52 player Vanilla coop co_52_zeu_benchmark_v001.Altis Params available are a) Select number of AI from 0 to 600 (In increments of 50) b) Viewdistance : (default 2.5k) c) Terrain Grid (Default = MP default of 25) d) Max allowed dead bodies (Default set to unlimited) e.g. no body removal e) Enable Simulation (Options to disable, None, All, All server or All client objects f) diag_captureSlowFrame option (Unable to test the functionalityu for this as the current versions do not support this command) g)Debug to server log (Enabled by default) ----------------------------------------------------------------------------- How should one test. TEST 1 This is to reproduce an issue where High player numbers and moderate AI numbers cause very low FPS serverside Test Requires a) Dedicated server with at least 100MB connection b) Minimum 30 players c) No addons Start mission with default params (dont touch any of them) The mission will load with 300 AI present on the map This should cause a very slow frame-rate serverside Run mission for say 5 mins max Your rpt will then have entriies similar to the following |========================= co_52_ZEU_Benchmark. Altis =========================| "" "" " _____INITIAL SETTINGS_____" " Viewdistance : 2500" " Terrain Grid : 25" " Count AllMissionObjects : 372" " EnableSimulation On all SERVER objects: true" " EnableSimulation On all CLIENT objects: true" " Max number of dead AI allowed on server: 0" "" "" "_____DATA OUTPUT_____" " Time: 1.191" " Players: 1" " Local AI: 308" " Dead AI on map: 0" " diag_FPS: 18.1612" "" "" You will get the following data (with differing values) repeated once every minute that you run the test mission "_____DATA OUTPUT_____" " Time: 1.191" " Players: 1" " Local AI: 308" " Dead AI on map: 0" " diag_FPS: 18.1612" "" "" This should easily reproduce the issue and once we can verify this over a few servers we can them submit a ticket for this issue with data we have collected If you do run this as described, please add the following in your feedback post 1) Server specs (CPU/RAM/Connection speed) 2) The .rpt file 3) Your ArmA3.cfg file Thanks Mission updated to v002 due to rpt log issue

I have finished the initial version of the benchmark mission This is a 52 player Vanilla coop co_52_zeu_benchmark_v001.Altis Params available are a) Select number of AI from 0 to 600 (In increments of 50) b) Viewdistance : (default 2.5k) c) Terrain Grid (Default = MP default of 25) d) Max allowed dead bodies (Default set to unlimited) e.g. no body removal e) Enable Simulation (Options to disable, None, All, All server or All client objects f) diag_captureSlowFrame option (Unable to test the functionalityu for this as the current versions do not support this command) g)Debug to server log (Enabled by default) ----------------------------------------------------------------------------- How should one test. TEST 1 This is to reproduce an issue where High player numbers and moderate AI numbers cause very low FPS serverside Test Requires a) Dedicated server with at least 100MB connection b) Minimum 30 players c) No addons Start mission with default params (dont touch any of them) The mission will load with 300 AI present on the map This should cause a very slow frame-rate serverside Run mission for say 5 mins max Your rpt will then have entriies similar to the following |========================= co_52_ZEU_Benchmark. Altis =========================| "" "" " _____INITIAL SETTINGS_____" " Viewdistance : 2500" " Terrain Grid : 25" " Count AllMissionObjects : 372" " EnableSimulation On all SERVER objects: true" " EnableSimulation On all CLIENT objects: true" " Max number of dead AI allowed on server: 0" "" "" "_____DATA OUTPUT_____" " Time: 1.191" " Players: 1" " Local AI: 308" " Dead AI on map: 0" " diag_FPS: 18.1612" "" "" You will get the following data (with differing values) repeated once every minute that you run the test mission "_____DATA OUTPUT_____" " Time: 1.191" " Players: 1" " Local AI: 308" " Dead AI on map: 0" " diag_FPS: 18.1612" "" "" This should easily reproduce the issue and once we can verify this over a few servers we can them submit a ticket for this issue with data we have collected If you do run this as described, please add the following in your feedback post 1) Server specs (CPU/RAM/Connection speed) 2) The .rpt file 3) Your ArmA3.cfg file Mission updated to v002 due to rpt log issue

semi related threads http://forums.bistudio.com/showthread.php?169097-What-performance-should-we-expect-from-the-following-hardware-mission-scenarios http://forums.bistudio.com/showthread.php?169098-Benchmark-mission-and-benchmarking-tools-request-to-help-fix-multiplayer and please vote for this ticket, if you want to help sort out MP http://feedback.arma3.com/view.php?id=12658 @ Dr. Hladik As if has been posted before Depending on server Any coop mission with 300 AI present from the start of the mission and 50 players connected to a server hosted remotely, at a data center will drag the server fps to below 5 Typically the actual numbers of AI/players depends on server specs etc and can vary from the following, as a rough example Players: 30 - 50 AI :75 to 200 The only missions that do not cause these issues are missions where the AI are created dynamically during the mission The issue isnt really seen with 20 players Our server old xeon x5570_can run something between 150 and 250 AI before the fps drops below 15 with around 20 players connected so why dont you make us a benchmark mission which we could all run for you and designed by you, with params that could disable various elements in the game or ensable various functiomns so we could do the testing and give you feeback.

Multiplayer in Arma 3 is suffering badly because of some major server-side and client-side performance issues. This thread is not really about the issues, but more about how we as a community can help to resolve them The following thread discusses the actual issues in more detail in some detail http://forums.bistudio.com/showthread.php?168337-Multiplayer-not-in-playable-state As a shortlist they are: a) Degradation in fps over a period of time b) Unable to run moderate numbers of AI when hosting for 30 plus players in a coop (70 to 150 ai dependant on server) c) A badly desynced player is able to drag the multiplayer session down to an unplayable level (Dyslexci's most disliked issue) d) Server admins do not really know how to configure their arma3.cfg (Bandwidth settings) correctly. I recently set out to create a coop benchmark mission that a large player base servers could run to help reproduce these issues. During creation of this benchmark mission, I realised, that I do not know what B.I would actually require in such a benchmark mission. What param options would be useful etc B.I know what data they need to collect and have internal tools that can collect the data. B.I should really be creating this test mission for us, if they want our help There is a very loyal fan base that is more than willing to help test and give feedback for these issues. There is a ticket raised for just such a thing (Please vote for it) http://feedback.arma3.com/view.php?id=12658 Kju has asked for these tools for many years now Could B.I please reply to this thread and help us to help you fix Multiplayer I am sure that we as a community could organise a large scale testing session across an array of servers providing we have 1) A benchmark mission 2) A set of tools to log required data 3) A set of instructions so all testing is carried out consistently What would also be useful, is if you could define what a server with a set spec should be able to run. Please see this post: http://forums.bistudio.com/showthread.php?169097-What-performance-should-we-expect-from-the-following-hardware-mission-scenarios

You should only need to log into steam on the server to update the package. Which doesnt happen that often, basically the same rate as the stable patches unless a hotfix has been created You don't need steam to launch or run the exe, just use a desktop shortcut or Firedeamon etc to launch the A3 server

i can add a dead object removal for testing, by default disabled and then with a benchmark fps you can rety nd see what difference it makes

s5till working on the benchmark mission, busy doing building works in the houses and every time i sit down to it i think... ah need a param that can disable That too

are you using different ports for steam, eg on the server don't use default

16:10 @ 1920 x 1200

You would need to ask Dwarden for a key, PM him

Player connection issue has been resolved. kudos to Dwarden for finding the solution change the steam_appid.txt content to 107410

@ KevsnoTrev ref the markers, (Nato group markers) they appear to be standard markers not groupIcons and only in map view You are thinking about the name tags which you can see when you have the hud enabled while wearing tactical glasses What i mean are _grp addGroupIcon << params >> _grp setgroupIconParams << params >> setGroupIconsVisible [true,true];//[select 0: show markers on map, select 1: show markers on player hud]; The beauty with these, is you dont need any looping script to setmarkerpos to group leaders BIS internal coding smoothly moves these markers with the group leader !!BUG!! On the latest rc7, reference the squad management gui On the drop down box that allows the admin to certify a player to be pilot, i cannot read the second entry in that list, hidden from view

re run the cmd file, it will verify all the files if that doesnt fix it, delete the folder at the top of the a3 install, the one that is just a bunch of numbers and letters and re run the batch file It is not uncommon for the patching to be corrupted

Dont worry about it just yet m8, the server package isnt synced with the stable client release yet (Some technical issue which means it has a delayed release following the stable release and you also cannot connect to the current dedi server package as a client (Bug)

The changelog in the first post states I am also waiting for the public release with eagerness. There is not a public co-op style mission out there that can compare to this. Even just from a server-side performance viewpoint, nothing compares. patrol ops has been around for each engine version and it always stands out from the crowd. This one though is on a different level, well done Roy and those supporting his efforts. I havent seen the latest version, am just wondering if the group marker system is using a) Group markers b) Group icons The latter is much smoother and gives you the option of a 3d marker in game and also clickable with whatever functionality you want to implement, which may be preferable especilly for Commander siutational awareness when played more seriously For example, typical functionality that could be employed. As mission commander GIVE MOVE ORDERS 1) click on marker (group icon) (group now selected) 2) click on map location (Where you want the group to move too) 3) remote Group then receives following information ....a) Localised marker positioned at their "Move to" waypoint ....b) localised to group, overlay text box opens that states they have received "Move" orders ....c) Waypoint appears on map, so they don't need to keep looking at the map Group Info Hud Maybe "mouse over" group icons opens up an information hud a) Group leader b) Manpower details etc This to me would be icing on the cake and help alleviate some of the language and non voice comms issues you get in a public MP environment This is just a wish list from me and am sure you get lots of them

for stable and if using the .cmd file defined in the initial post, use SET A3BRANCH=107410

I have no idea how dead objects impact performance. I was quoting Dr. Hladik statement I would assume, that code being good, the dead objects should impact less than alive objects, for such obvious reasons as no expected state change etc. However, the quote was in response to fps degradation over a period of playing time and the assumption was made that it could be caused by increasing numbers of dead objects. If anything, with less alive objects on the map and no new alive objects introduced, you would expect FPS to increase. However this is not what folks are stating. I will add an optional clean up system to the params of the benchmark mission, that way we can get a comparison in a gameplay environment

Why would a dead object, not moving, not changing state create more processing-data than something that is alive. That must be an engine issue If it is a requirement of a mission to delete such objects to maintain a playable state, then the engine should deal with it automatically. It's unfair to expect a mission designer (Especially not a time served one) to be aware of these issues, even less so to know how to implement such a function unless there is a mission design document that states how to do it correctly. A Description.ext switch would be far better than a module in this instance Deletedeadobjects = TRUE; TimeToDeleteDeadobjects = 600; // Time delay //or DeadObjectLimit = 50; // Dead object Array size limit when count is 51, select 0 is deleted (Please bare with me for benchmark-test mission, finding time is troulbesome at the ,moment, 75% done)

so what is the server fps at this point?