TERA 100.02 VM

[babyshaq10]

yeah thats right when i start the server everything is so Lagg, thanks man for this info.

 

[Messmeriner]

Does anyone know what the dungeon other in the server manager adds or does?

 

[Archgeus]

Does anyone know what the dungeon other in the server manager adds or does?

Check ServerConfig > WorldServerList,

Those separated world servers are exclusive for certain continents

Ex. World server with id 15 is used for event continents like Manglemire in lower versions for example

 

[lolidroid86]

I have an issue... i was running the server fine on a dell r720 server dual 10c/20t with 128gb of ram from an nvme on a pcie adapter, then due to electric use i moved the server vm from that server to a dell t7500 with dual 6c/12t with 128gb of ram and the vm on a pcie adapter, now i had to downgrade form vmware 26h1 to vmware 16.2.5 due to the xsave instruction set missing on the t7500, when i was on the r720 i had no problem running the vm i was running the full server and played for 2 days no problems but like i said due to power usage the r720 uses dual 1000 watt power supplies and the t7500 uses a single 1000 watt power supply, i changed over to it, nut now every time i try to load the world server i hit a null exception when the world server starts checking for monster movement, by like the 3rd check the world server crashes out with the mentioned null pointer at address 0x00007FF797B7BDB6, now im trying to debug the problem with claude code and were now trying to debud the minidump created with the crash, but not having alot of success, is there a way you could share the acutal source code for the server not the precomplied stuff but the barebones of it so we can look into trying to fix the actual null pointer error in world server or is that not possible. any help here would be truely appreiated.

note: its an access-violation null-pointer write, not an illegal-instruction crash

update 1: crash is in Territory::CreateNpc, null pointer written after what looks like a failed pool allocation

 

[Archgeus]

Source code doesn't exist publicly for TERA. Publishers only had access to precompiled binaries. The only one which had the source code was bhs itself, but i doubt that they had the src alongside the live/QA binaries.

 

[lolidroid86]

is there anything you could point me towards to maybe help me further?

 

[dziobus]

is there anything you could point me towards to maybe help me further?

If it crashes on a file check, then it is safe to assume there might be some corruption in the files it gets stuck at. Have you tried simply replacing those files with new ones?

 

[wanqwer]

i was running the server fine on a dell r720 server dual 10c/20t with 128gb of ram from an nvme on a pcie adapter, then due to electric use i moved the server vm from that server to a dell t7500 with dual 6c/12t with 128gb of ram

Sorry that this is off topic but since I'm more of a hardware geek - I had to ask:
1. Do you know that those 2 power supplies are only for redundancy and only uses one at the time?
2. r720 is an older generation than t7500, so you should not get electric efficiency upgrade. Could have just set it to "power efficiency" on ram and cpu in bios and you would have been golden.
Just my two cents.

As for your error, there is no reason for that error to appear if you just copied the VM from one physical disk to another.
If you can, transfer the original hdd/ssd you used to the new machine and try again - and if that works - just copy everything again, maybe the new drive has bad sectors.
Another tip - if you copy not the VM but only the server files to another place/disc, delete all log/dumps/crash files that were previously generated, don't know why, but I had problems with them before.

Can use something like this from root server directory, saved as .cmd

@echo off
echo Removing .log and .logs...
for /r %%F in (*.log, *.log2, *.logs, *.crash, *.dmp) do (
echo Deleting: "%%F"
del /f /q "%%F"
)
echo Done.
pause

 

[CHILAKIL02]

GUYS, the matchserver is working on this version?

i've a problem when i open the client the splash window never finish and the client no start, any help?

 

[trifecta]

Has anyone figured out how to prevent bosses from resetting to 100% hp upon death / resurrection? I intend to play mostly as a solo dungeon player so wipes are inevitable

 

[outlanderzor]

Has anyone figured out how to prevent bosses from resetting to 100% hp upon death / resurrection? I intend to play mostly as a solo dungeon player so wipes are inevitable

I believe you can achieve this by setting hpResetWhenCombatEnd to false on the <CombatState> element in AIData.
Same config is used for Kezzel in v31.

 

[andrea89]

Hi Archgeus, first of all thank you for sharing this VM.


I’m trying to run your TERA 100.02 VM locally for solo play, but I have a problem with WorldServer and I hope you can help me.


My PC has an Intel i9-13900F, 64 GB of physical RAM and NVMe SSD. I’m using VMware Workstation 17.6.4. I assigned about 58 GB RAM and currently 16 vCPUs to the VM.


I am testing a completely fresh extraction of your VM. I haven’t modified ContinentData.xml or any other datasheets.


I start the services manually and I keep LogServer disabled. ArbiterServer and TopographyServer start correctly.


The problem is WorldServer. It keeps loading for many hours but never reaches $ Server ready $.


During loading it periodically prints many messages such as Npc act delayed in anger mode and also NpcArenaStart. Last night I left it running all night. It continued printing new lines until about 5:35 AM and then stopped producing new output, but $ Server ready $ never appeared. WorldServer.exe was still running and did not appear to crash.


I checked Task Manager while it was loading:

• WorldServer.exe: about 31.8 GB RAM
• TopographyServer.exe: about 4.5 GB RAM
• ArbiterServer_m1.exe: about 3.3 GB RAM

The VM therefore gets very close to using all the available physical memory.


I know you previously recommended 32 GB physical RAM + 100 GB swap on SSD, or 128 GB RAM for full functionality, so I’m confused why my fresh VM with 58 GB assigned cannot finish loading WorldServer.


Is this extremely long WorldServer startup normal with 64 GB physical RAM, or does it sound like something is wrong with my configuration?


Is there a specific RAM/pagefile/vCPU configuration you recommend for your 100.02 VM when it is only being used by one player locally?


Also, should WorldServer always eventually display $ Server ready $ before I try to enter the game?


If necessary I can upload my WorldServer log and screenshots.


Thank you very much for your help.

 

[bbz]

It throws an error

 

[trifecta]

Has anyone figured out how to add npc ai or pvp ai bots for dungeons or fwc / cs? For example, spawning a tank / healer.

I see some ai files so thought it might be possible.

 

[outlanderzor]

Has anyone figured out how to add npc ai or pvp ai bots for dungeons or fwc / cs? For example, spawning a tank / healer.

I see some ai files so thought it might be possible.

Technically feasible, but not trivial if done manually.

You can get tank/healer NPCs from Ravenous Gorge and transplant them into other dungeons, but that requires tracking down all their dependencies, which are spread across multiple files: NpcData, AiData, NpcSkillData, TerritoryData, DungeonData, and String files.
You must ensure that the IDs of the entities you're transplanting into the region don't clash within the same huntingZoneId. Once you have everything ready on the server, you also have to sync the modified server files back to the client.

 

[eronaong]

Successful implementation of web store in Korean client

How did you implement the web store in the KR DC version? I would appreciate it if you could share the method.

 

[KoriKo]

I left the server running while I slept, and when I returned, the PC had rebooted and a password had been set.

I suspect a Windows update download caused the reboot, but I cannot figure out why a password was set.

 

[trifecta]

Could anyone confirm whether the battlegrounds work? I've started battlefieldserver and matchserver, and set the requirements to enter from [5] to 1 per side, but when I attempt matching on two different characters nothing happens. Anyone experience this?

 

[trifecta]

To add more context to my previous post, every matching attempt ends with worldID[-1]. Both players are successfully matched, but MatchServer then appears unable to find an eligible battlefield host.

To test Skyring which is the easiest, BattlefieldServer initially registers as worldId[10], and MatchServer lists instanceId[38] count[0], but the logs also show worldId[10] disconnecting from ArbiterServer. Has anyone encountered this before or know what could cause the BattlefieldServer–Arbiter connection to drop? Is there a configuration change or workaround that keeps it registered?