[CRITICAL] Connection TIMED OUT.

ChillyRider · April 15, 2026, 4:40am

Summary

Our team is struggling with this issue and can’t find a solution.

Previously, we had around 60k actors and 94k memory usage, and we were at least sometimes able to launch a session. After reworking our logic, optimizing code, reducing actor count, and even commenting out everything in OnBegin in Verse, we are now completely unable to launch a session.

Current state:

~45k actors
Spatial loading is enabled
~80k memory in the center
~90k at the highest point

We don’t understand how this is possible. I’ve spent over 10 hours trying to launch a session, and nothing helps.

At this point, we would really appreciate assistance from a technician to determine whether this is an issue on our side or something else.

project id: 7bc8cb13-45aa-75b2-733e-bfaaa021be79

Please select what you are reporting on:

Unreal Editor for Fortnite

What Type of Bug are you experiencing?

Other

Steps to Reproduce

Attempt to launch the latest build.

Expected Result

Session launches successfully without timing out.

Observed Result

Session always times out, regardless of changes made.

Platform(s)

PC

Upload an image

Logs.7z (5.09 MB)

ChillyRider · April 15, 2026, 4:09pm

I saw that Epic had server issues and marked them as fixed 4 hours ago, but nothing has been fixed for our map. It’s impossible to launch a session, test, debug, and so on. Our workflow is completely stalled.

ChillyRider · April 16, 2026, 1:20am

29k actors. 73k memory. 2 hours of trying lunch session… Nothing.
I need to delete hole project to start session?

Jimbohalo10 · April 16, 2026, 10:33am

Hi @ChillyRider

Mine is NOT EPIC just an Indie Developer and no warranty so its your responsibility NOT MINE OR EPIC is given to the stuff below.

DONT DELETE ANYTHING

This is what is in every log you supplied

This is the error stopping you launch

[2026.04.15-00.27.04:155][432]LogValkyrieRequestManagerEditor: Error: Failed to handle connection attempt (

Network error occurred when connecting to a server after matchmaking completed.

here is the bits working

LogValkyrieSummary: Project Tiki up to date

LogValkyrie: Play succeeded

Now the this is start of failure trying to get Get Project Cooked Size

LogValkyrie: ProjectSizeRequest::GetProjectCookedSize
[2026.04.15-00.27.04:155][432]LogValkyrieRequestManagerEditor: Error: Failed to handle connection attempt (Network error occurred when connecting to a server after matchmaking completed.

Here is the status report on the size of Project

LogInstallBundleManager: Display:
[2026.04.14-23.06.34:531][ 88]LogInstallBundleManager: Display: * Install Bundle Cache Stats ContentOnDemand
[2026.04.14-23.06.34:531][ 88]LogInstallBundleManager: Display: * size: 5368709120
[2026.04.14-23.06.34:531][ 88]LogInstallBundleManager: Display: * used: 0
[2026.04.14-23.06.34:531][ 88]LogInstallBundleManager: Display: * reserved: 0
[2026.04.14-23.06.34:531][ 88]LogInstallBundleManager: Display: * free: 5368709120

5 GB streaming will take a while to load!, this would take 17 minutes to load on my laptop with Wi-Fi 6

Something to do IMHO

Before deleting anything

Now 40.20 is available go to ANOTHER PC do a fresh installation of Fortnite/UEFN.
Restore the project using Unreal Revision Control “sync” function to put a clean version of project on PC use URC selector in the new version where you can select BEFORE the problem occurred.

Cant find the new documentation for URC 2, but its in a posting somewhere and maybe there is documentation.

Next try using an Alternate Epic account in the same team make a new account in the Team

The logs show all your assets are successful validation.

Backup the main project with Zip/Rar and call it something and try loading it onto the new PC.
Check your firewall settings and network settings in past a provider has changed the amount of usage called throttling.

Try to publish a private island get a private island code. set “show in Creative account” using the UEFN Publish function. See what errors you get Memory of Assets will be checked.

Ultimately publishing the island is what development is about

Are you on a 1GB fibre unlimited usage?

WAIT FOR EPIC SUPPORT before deleting anything

if you do delete make a backup first and delete stuff on the backup first

Bug-Reporter · April 16, 2026, 9:00pm

The status of FORT-1088719 changed to ‘Needs Triage’. We’ve confirmed the issue and it’s waiting to be assigned to someone to fix it.

ChillyRider · April 17, 2026, 5:44pm

During the night we were able to launch sessions a few times. However, the next day, after 4 hours of attempts, not a single session launched successfully.

@Flak Apologies for the repeated ping, but we urgently need a fix as the Easter event is nearing its end across all maps.

ChillyRider · April 18, 2026, 4:53am

For the past three days, I’ve been able to consistently reproduce successful session launches in UEFN only during a very specific time window — roughly between 5:00 AM and 7:00 AM (GMT+2).

This leads to a few conclusions:

Server capacity appears to be overloaded, and there seems to be no distinction between players and developers. Developers are clearly not being prioritized, which makes it extremely difficult to work.
Our map is within normal limits (around 80k memory temps, project size ~63%), and we can occasionally launch and test sessions during off-peak hours. This strongly indicates that the issue is not on our side, but rather something only Epic can address.

I’m not sure if this is a Europe-specific issue or a global one, but it is quite serious. It needs a permanent fix, not temporary patches every few months. Right now, it feels like each update introduces instability again, and developers are left dealing with recurring problems.

We urgently need a stable and reliable environment to continue development.

Jimbohalo10 · April 18, 2026, 9:49am

hi @ChillyRider ,

There is another user who supplied a failure log for me to look through.

Here is the most detailed report for Quality Of Service ever seen.

In the end ASIA was picked as the best area but ping times are not great IMHO these are other problems


[2026.04.17-13.22.02:528][768]LogEOSSDK: LogEOSAuth: UserAuthGenerated: Received FUserAuthToken
[2026.04.17-13.22.02:589][788]LogPing: CheckAwaitReplies: 13.37.152.32:22222 (id: 38, seq: 37) timed out after 3.0000 seconds
[2026.04.17-13.22.02:591][789]LogQos: UQosEvaluator OnEchoManyCompleted; result status=1
[2026.04.17-13.22.02:591][789]LogQos: Verbose: Qos complete in 3.40 s  (all regions: yes)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: ping average detail per datacenter:
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	VA (NAE): 3/4 queries succeeded, average ping: 221ms (adj: 221ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	OH (NAE): 4/4 queries succeeded, average ping: 218ms (adj: 218ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	TX (NAC): 3/4 queries succeeded, average ping: 204ms (adj: 204ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	IA (NAC): 3/4 queries succeeded, average ping: 268ms (adj: 268ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	MX (NAC): 2/4 queries succeeded, average ping: 209ms (adj: 209ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	NCAL (NAW): 3/4 queries succeeded, average ping: 156ms (adj: 156ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	OR (NAW): 4/4 queries succeeded, average ping: 183ms (adj: 183ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	DE (EU): 4/4 queries succeeded, average ping: 196ms (adj: 196ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	GB (EU): 4/4 queries succeeded, average ping: 213ms (adj: 213ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	FR (EU): 3/4 queries succeeded, average ping: 203ms (adj: 203ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	SYD (OCE): 4/4 queries succeeded, average ping: 185ms (adj: 185ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	SAO (BR): 4/4 queries succeeded, average ping: 333ms (adj: 333ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	TOK (ASIA): 4/4 queries succeeded, average ping: 56ms (adj: 56ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	BAH (ME): 4/4 queries succeeded, average ping: 154ms (adj: 154ms)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: [UQosRegionManager::OnQosEvaluationComplete] bSubspaceBiasEnabled=true
[2026.04.17-13.22.02:591][789]LogQos: Warning: No datacenters for region UND
[2026.04.17-13.22.02:591][789]LogQos: [UQosRegionManager::RefreshUsableSubregions] Refreshed subreion list. [TOK,BAH,NCAL,OR,SYD,DE,FR,TX,MX,GB,OH,VA,IA,SAO]
[2026.04.17-13.22.02:591][789]LogQos: Verbose: No region currently set.
[2026.04.17-13.22.02:591][789]LogQos: Verbose: [UQosRegionManager::TrySetDefaultRegion] setting default from GetRegionId() (NONE)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: [UQosRegionManager::GetBestRegion] Best region: "ASIA"  (Current selected: "")
[2026.04.17-13.22.02:591][789]LogQos: Verbose: [UQosRegionManager::TrySetDefaultRegion] setting default from best (ASIA)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: [UQosRegionManager::SetSelectedRegion] Old: ""  New: "ASIA"  (force? false)
[2026.04.17-13.22.02:591][789]LogQos: Verbose: [UQosRegionManager::GetBestRegion] Best region: "ASIA"  (Current selected: "ASIA")
[2026.04.17-13.22.02:591][789]LogQos: [UQosRegionManager::OnQosEvaluationComplete] ping eval has completed.  Best region is 'ASIA', recommended subregion is 'TOK'
[2026.04.17-13.22.02:591][789]LogQos: Verbose: UQosEvaluator OnEchoManyCompleted; Qos result=1
[2026.04.17-1

Used scripting highlight

[2026.04.17-13.22.02:591][789]LogQos: Verbose: 	MX (NAC): 2/4 queries succeeded, average ping: 209ms (adj: 209ms)

n

NA Central fails half time only 2 out of 4 queries are accepted and is worst

France has 3/4 means drops every 4th request

GB London,mine, is fine 4/4 bearing in mind he is in ASIA.

ASIA was chosen because of the lowest ping rate 56ms

This is NOT about matchmaking its general networking basic PING data by UEFN own software which backs up your previous post and my original answer.

ChillyRider · April 19, 2026, 3:16am

Day 4 of consistent and successful session launches from 5 AM to approximately 7 AM. During the day, I switched VPN to Japan and was able to start a few sessions to debug the map. Same memory, project size, actors on the map.

@Flak How long do we have to struggle?

danger-dev · April 19, 2026, 7:25am

thought I was going crazy, guess this is an issue on EPICS end… makes trying to fix bugs or push an update impossible.

ChillyRider · April 19, 2026, 7:31pm

From my perspective, things are getting out of control. Our team has been working relentlessly for almost a year, only for Epic’s incompetence and lack of organization to undermine everything we’ve done. The Easter event is now over, and we’ve received no response, no ETA for a fix, and no communication whatsoever.

ChillyRider · April 20, 2026, 4:22am

Day 5 of consistent and successful session launches from 5 AM to approximately 7 AM.

@Flak ?

ChillyRider · April 20, 2026, 7:38pm

Just out of curiosity — what are the actual current limits for developers when it comes to memory, actor count, and project size?

Right now it feels like you’re effectively capped at around 50k memory, ~20k actors at best, and less than 50% project size if you want any realistic chance to launch and debug a map consistently.

Is this the intended baseline, or are these just side effects of the current backend/server instability? Because in practice, anything beyond these thresholds becomes unreliable or completely unusable.

Jimbohalo10 · April 20, 2026, 7:45pm

hi @ChillyRider

Please could you add a 64GB Virtual Memory (pagefile) to Windows 11

Just to see if this makes any difference you may need 128GB (128000)

How to add 64GB Virtual Memory pagefile

ChillyRider · April 20, 2026, 7:47pm

Already done. 64gb page file and 32gb Ram.

ChillyRider · April 20, 2026, 10:16pm

New update arrived. 0/5 successful session launches. Switching to Japan still occasionally helps to launch a session — maybe 1–2 successes per hour.

ChillyRider · April 21, 2026, 4:32am

Day 6. Only 1 session launched between 5 AM and 8 AM. Getting worse?

@Flak ?

Flak · April 21, 2026, 11:24am

Any updates will be posted by bug-reporter. I’m afraid there is nothing new to report here.

ChillyRider · April 23, 2026, 6:54pm

To test the map, we’re forced to upload a private version every single time. Over the past few days, gameplay has been plagued with ping spikes, lag, and packet loss (this is not on our side — we tested multiple ISPs across different cities). Now it’s getting worse with connection losses and random disconnects. If fixing server stability isn’t priority #1, then I honestly don’t know what is.

The_Creator_001 · April 23, 2026, 7:31pm

Hey there, based on my issue which is similar to yours, the problem you are encountering could be not with memory caps on amount of devices you are using or actors, but due to device bindings, there must be some new limit due to the new way of uploading maps.