Pve-c 1516 bad lag and spikes

Game mode: OFFICIAL SERVER
Type of issue: PERFORMANCE
Server type: PVE - Conflict
Region: AMERICA


Server has been extremely laggy now for 2 weeks. I had posted earlier and was asked to run the WinMTR program. I have ran the program a few times now and sent the results to Funcom and to GPortal. It has been 4 days now with no feedback on the information I have provided nor any improvements on the server. The logs clearly show that there is issues with connections to the server that is causing the lag. The servers population has now dropped in half. Could we PLEASE get some feedback as to the issue and if there are steps being done to correct it before we are go from being one of the most populated to a dead server. I have added my recent WinMTR log readout as well. Thanks in advance.

04172020

I am no expert but doesn’t this show that the package losses and high ping start right outside your “door” and remain more or less consistent all the way to the server destination?

1 Like

More the other way around. I know it isn’t my side especially as it is the whole server suffering

Can confirm, bad spiking in peak hours, around 15-20 players online.

If the lag is caused at the server end, by that trace it’s not to do with their network connection, got to be something else. Breaking down the trace you show, the only (minimal) packet loss is at your service provider’s end:

image

As an aside, while I have no experience with WinMTR myself, I would question the loss displayed being a percent loss, the numbers don’t add up…

Edit - sorry, missed one out:
image

With all of the hops though, the average ping builds as I would expect (higher the further they get away from your home site). Looking at it I don’t think that any issue you are seeing relates to network latency at either end of the connection…

They have stated that it is an issue on the GPortal side. But this was done to hopefully help them narrow it down and fix the issue. The server is still waiting for this fix after a month now. This happened for a period of 2 weeks. Then things seemed somewhat normal for 2 weeks. Now we have been back to the trash for another 2 weeks and still lagging horribly. Players are not only quitting the server but the game as well. It is absolutely ludacris that it still going on after a month. I love that Funcom is still adding and improving the game but what good does it do the players if we can’t play the game due to faulty servers?

i dont need a report to see that my ping has not gone once under 500 in the past 2 weeks. even when there was 3 people on the server. 1517 had the same fate 2 weeks ago.

if oyu hop on different servers in NA you will notice that not all of them are under the same stress

Can confirm what others are saying. I joined the server mere days before these issues began and didnt notice anything amiss until logging in one night to 1020 ping constantly for minutes on end. this is now going on 2 weeks of the same. please for the love of god can we get some maintenance on this server?

I wouldn’t challenge your experience; I was just analysing the detail of the network trace. Like @DrakonTrelos notes above your post, Funcom have confirmed to him that the problem resides at the server destination.

As I said above, my only intent in posting was to note that the network trace looked fine, with the only (tiny) packet loss at the origin network rather than at the destination. Based on your posts before (and the posts of others) I would have to be an idiot to argue there wasn’t a problem somewhere.

Just not the network :slight_smile:

It does suck, especially as you are stuck in the middle of a problem that exists outside of your control.

Trouble is, having also been at the service provider end of this sort of problem, it’s a total nightmare to troubleshoot. Some problems only display themselves when services are actually really needed and under full load. They then disappear like mist when you look at them outside of the crisis point.

I’d like to add to this thread as well as a new player on #1516. Something isn’t right, why are some servers fine and this one and a few others not? And for those saying it looks like it’s client side it’s not, everyone is experiencing the lag and it’s been going on for two weeks straight now. It’s horrible and most times unplayable. People are afraid to run dungeons for fear of lagging out and dying, which has now happened to me personally. I get that it’s not a paid service but we do buy your DLC, I have all of it but I’m. It going to spend any more money until the server that I have invested a lot of time on is fixed. Please address this. You shouldn’t need any more info from us, just have your techs log in and see for yourself how absolutely dreadful the lag is.

No one in this thread has said it is client side. Funcom hasn’t said it is client side when responding to DrakonTrelos. Who has said that this particular problem is client side?

Was disconnected this morning 11:14am Central time 4/21/2020. Also I was the only one on the server early this morning and was still getting large ping spikes of the dreaded 1020 ping. Any word from funcom or Gportal?

Unfortunately there has been no word from Funcom or GPortal on any progress in correcting the server. Almost at the point where if I do quit I won’t be back until the servers are reset for everyone with a fresh start. Been on this server for almost 2 years and just not up for starting over unless it’s a level playing field

Server just crashed for everyone at 4/21/2020 at 11:34 central time What is the plan here guys what can we do??? Should we post at G-portal as well? Looking for some kind of guidance??!?!

Well I will take it that because I have not heard from a Dev. Or a community support representative that there must be no answer for this POOR performance coming from the GPortal server. I will pass the word along to the players on the server that there has been no response in regards to the horrible ping issues and leave it at that. Too bad that our voices have not been heard and it is sad that we are now losing even more players and servers now dying due to this poor performance. This was a great game while it lasted though so thanks for the past 2 years.

They are coasting until the expansion and milking us to buy DLC content. The Dune game is the focus more than likely, but hopefully they drop g-portal as the hosting provider or provide actual support for official servers. If not, no chance I play the Dune game whatsoever.

Based on the performance the server is likely on an oversubscribed host. If they could move us to another host they would have already instead of ignoring the thread.

I have fiber at my house and regularly check BGP looking glasses to check my providers peers to check for performance issues. I never find anything, but when I looked up the BGP ASN’s for G-Portal and their internet providers I have found occasional peering issues, but not constant.

The problem is likely tied to a contract with g-portal providing X number of servers for a specific amount of time. I played during EA, so surely a good portion of the servers are getting close to being EOL/EOS. I can’t think of any other plausible reason a new host would not be provided for the clusters of official servers complaining of issues.

Granted, pve-c and pve servers run like crap since people build massive compounds, so maybe funcom fixing their map design problems is one of the reasons they haven’t responded.

It could also be that the Unreal Engine libraries in the game folder are ancient and not even running a supported release version unless they forked off the main release. I did a binary compare of the library files against the ones on the Unreal Engine site and they match from a SHA-512 hash perspective. The version of Unreal Engine is a couple of years old, maybe three and behind a ton of patches. That explains to me personally why some things I reported as bugs in EA were never fixed, but show up as patched in the Engine patch notes.

Maybe someone will chime in with something to fix the issue or tell us to wait to pay for an expansion so they can generate revenue while they update the Unreal Engine and hopefully move to instanced dungeons instead of reserved hidden locations on the map. I would have loved to been in that Agile meeting, lol.

Conan Exiles GA Release Date / May 8, 2018

Install location -> \steam\steamapps\common\Conan Exiles\Engine\Build\Build.version\

Unreal Engine 4.15.3
{
“MajorVersion”: 4,
“MinorVersion”: 15,
“PatchVersion”: 3,
“Changelist”: 0,
“CompatibleChangelist”: 3299760,
“IsLicenseeVersion”: 0,
“IsPromotedBuild”: 1,
“BranchName”: “”
}

4.15 mainline / Released February 15, 2017
4.15.3 Hotfix Released 4.15.3 Hotfix Released / 05-23-2017, 09:33 AM

The number of fixed bugs and issues since 4.15.3, several thousand. The game launched on an old engine since updating to the latest engine before launch wasn’t possible. Explains a bunch of things to be quite honest.

Can someone from funcom provide an answer to the original question? Is the problem tied to pve/pve-c servers overwhelming a host due to a likely massive amount of entities? Is this a g-portal host issue?

Last I connected with three people online and had a constant 1020ms response time, but when using a tcpping to the tcp socket the three-way handshake was completed within 80-90ms. Is it a vswitch config issue, lacp, etc? Seems very localized to the handful of official servers complaining. I have spot checked several of the other posts complaining about the same thing on other servers and see the same results from my troubleshooting.

I am a global network, network security and datacenter architect, so ping me if you would like me to figure it out. Should I open a case with g-portal? Since these are tagged as “official servers” you would be the customer of record, so they will likely not tell me the time of day.

I can provide BGP peering information for g-portal, myself, and packet captures that show the problem. At first I thought it was a host being brought up with jumbo-frames enabled, but I don’t see the do-not-frag in my packet captures and the MTU from what I can see on the outside is what I would expect with the tier two/three carriers that provide their service.

Any feedback would be greatly appreciated since I enjoy the server and it sucks to login every few days to refresh a base with a ping over a second that I could likely troubleshoot in less than an hour on a webex.