Crash, Crash, Punch II

GTA V version? 1.47
Up to date? Yes
Legit or Pirate copy? Legit
Steam/CD/Social Club? Social Club
Windows version? 8.1 Pro (7 Fast, 10 Secure through 2023)
Did you try to delete caches.xml and try again? Yes
Error screenshot (if any) None
System specifications Intel i7-4790/MSI z97 14GB/GTX 760 2GB/Drivers Up to date Intel i3-3225/Intel HD61CR 8GB/GTX 560 1GB/Drivers Up to date
What did you do to get this issue? Played FiveM
What server did you get this issue on? None/My Favorite

Previously on this old server: Crash, Crash, Punch This server has run on FiveM since 2017, never one-time crashed in 2 years, as it ran 24/7 on Debian 9 updated, with 512MB RAM, while running apache2 with 1 busy website. Both server and website each get about 10k users a month. Both ran without comment or issue of any kind for two years. The FiveM elements, asserted this was my fault, that apache was using all my RAM, and that it had nothing to do with the forced 1317 linux server upgrade for FiveM, which on that day of upgrade, my crashes began.

Before 1317 my server had never crashed, and with no changes to it in 6 months time after 1317 my server crashed every day for two months and destroyed my community, from averaging 20 users at any point in the day to less than 5 as the server crashed again, and again, and again, and again. All crashes where FiveM server Killed. All OOM. (Out of Memory)

I rebuilt this reasonably new VPS server that had been perfectly stable for 2 years prior to 1317. Did not install Apache. FiveM only, and it crashed, and it crashed, and it crashedā€¦ OOM! I moved to a new VPS, that had 2GB of ram and it never crashed in 2 weeks. I moved it back to a new build of my old reliable 512MB VPS, with no apache, nothing but FiveM, everything up to dateā€¦ and it crashed, and crashed, as i tried FiveM server upgrade after upgrade. Where my old VPS has all the power I need, the new VPS was barely capable of running FiveM, everything took 10 times as long to process, but due to tons of ram, it never crashed.

Where the workaround ā€˜buy more RAMā€™ stands, the real question remains, why is post-1317 using so much more RAM than before, or more erratically? Apache, is not now, nor was it ever the cause of these crashes as they continue with no Apache on the VPS. FiveM linux post-1317 is on a death march to OOM, for everyone. The average FiveM Server Crash now occurs between 60 and 72 hours uptime, where it once was ever 12-24 hours.

And here are new logs: messages and syslog from old crashy, running Linux DebianUW2 4.9.0-8-amd64 #1 SMP Debian 4.9.110-3+deb9u6 (2018-10-08) x86_64

1908 - 1375 - Crash crash punch - syslog and messages.zip (123.6 KB)

Please answer for me:

  1. Why is FiveM using more RAM since 1317? And why not before?
  2. What else can I provide you with to help solve this problem?

Hmmmā€¦ Canā€™t add bloatware as optional tag in support forum topic! This could be a problemā€¦

Iā€™m pretty sure the consensus was you needed more RAM to support your server. Try not running it on a tiny VPS with other memory-hungry services instead of being stubborn and looking for answers that donā€™t exist.

Just my two centersā€¦to be fair, if youā€™re going to quote a post - quote the whole actual post please.
Crash, Crash, Punch or something like thatā€¦

1 Like

Things change during updates, its how it is so over time of course the more thats added to it its gunna use more memory.

Your solution is to not use a low end VPS.

This consensus you speak of, was based on the false claim made by multiple elements that Apache was using all my RAM. And that FiveM was only getting 128MB of ram. This is not the case now. FiveM cannot contain itself to 512MB of ram due to your 1317 upgrade being unstable.

This was my original point, that the elites of FiveM and their 128 slot OneSyncs with 200 players in queue, were by the 1317 update adding ungainly bloatware for us everyday Joeā€™s and our 15 players who donā€™t want to spend $250 a year to run a server 24/7. But thank you for noting, that this problem is caused by the update.

My ā€œCheapā€ VPSs have at their peak run FiveM, JC3MP and 2 websites simultaneously on 512MB ram, while JC3 was topping 80 players 10 hours a day. And on a second run MTA (35+ players), ā– ā– ā– ā– ā– ā– , and 4 websites on 512MB ram, for more than a year in both cases, without any problems. And Rage was running in ram heavy Docker at that. Without any problems at all. I have beaten these VPSs like a rented mule year after year and they always perform as expected. Conversely the 2GB VPS I switched to for 2 weeks was so bad I had to leave it as soon as I could. .tars that take my cheap VPS 10 seconds took it 2 minutes. But i digress, FiveM ran perfectly for 2 years on 512MB of ram, until the same day the 1317 update was forced upon us, with no possibility of remaining on the previous stable builds, and with that you force the common man out of FiveM, to serve your elitesā€¦

And to what end? FiveM server is leaking ram. So how much more are you costing everyone by an inefficient forced update? And when you finally look back at this, will you all realize how stupid it is to categorically deny existing problems, as you make excuses for bad code, bad ideas, and bad management, where you should be working harder to fix problems. Anyone who has been Online knows this multiplayer has its limitations. IV was broken. V is broken. But FiveM worked, and there was nothing wrong with 32 slot FiveM.

My VPS Server notes, where I try to record things that happened, to look for patterns when things go wrong. It is not complete, it is just what i remembered to record. --[I inserted Server package release dates as a reference.]ā€“

19/01/27-02/10 fiveM 6% cpu 36% ram <---------------- 15 day uptime. No RAM leakage.

19/02/18-02/27 FiveM Processes CPU 7.6% MEM 44.2% <-- 10 day uptime. No RAM leakage?

19/03/08 Rebuilt FiveM from 1046 to 1056

ā€“[2019-Mar-19 07:20 1141]ā€“

19/04/10 ~9:02pm FiveM Server Disconnected 10 users, VPS server did not crash uptime 9 days, fivem server did crash or rebooted <-- An isolated Server Crash on 1056. Would not repeat for 2 months. Perhaps unrelated.

19/05/02 UPGRADED FIVEM to 1141

ā€“[2019-May-13 09:55 1244]ā€“

19/05/23 UPGRADED FIVEM to 1244

ā€“[2019-Jun-09 12:50 1317]ā€“

19/06/11 FiveM server crashed, rebuilt to 1323.

ā€¦it is all crashes after this point, where FiveM linux server went on to crash 50+ times over the next month.

190818

DID YOU LOCK MY POST AFTER 3 DAYS TO HELP USERS ENJOY A BETTER MORE STABLE SYSTEM? Or did you lock it because this is a known flaw, that you have no reason to help solve, and no answer to?