Roy
Content Type
Profiles
Articles
Updates
Projects
Twitch
Website Bugs/Suggestions
Guides
Newsletters
About GFL
Knowledge Base
Expenses/Hardware
Server Comparisons
Routing
Form Bugs
Community Representative Applications
Development Request
Forums
Events
Posts posted by Roy
-
-
The hop with NTT via Vultr has improved highly the last two days. I am finally getting the latency/ping I should be receiving to the PoP without any packet loss. I hope this keeps up!
Vultr is still investigating the Telia situation on the Dallas PoP. I will be requesting an update in the next couple of days.
Thank you.
-
Dealing with this through PMs, I will post back the end-result when completed. This isn't a firewall issue since the user can ping the server and receive a reply without any issues.
Thanks.
-
Hi all,
I just wanted to post this KB thread for any other server owners experiencing this issue. A while ago, our TF2 servers (2Fort and High Tower) were experiencing a crash after 2 - 3 days of up-time. The stack trace on Linux would look something like this:
#0 0xf6725fe1 in Sys_Error_Internal(bool, char const*, char*) () from /home/container/bin/engine_srv.so [Current thread is 1 (Thread 0xf7d6d700 (LWP 48))] #0 0xf6725fe1 in Sys_Error_Internal(bool, char const*, char*) () from /home/container/bin/engine_srv.so #1 0xf6726030 in Sys_Error(char const*, ...) () from /home/container/bin/engine_srv.so #2 0xf6727348 in Sys_SpewFunc(SpewType_t, char const*) () from /home/container/bin/engine_srv.so #3 0xf7d4b0ab in _SpewMessage(SpewType_t, char const*, int, Color const*, char const*, char*) () from bin/libtier0_srv.so #4 0xf7d4b7cc in Error () from bin/libtier0_srv.so #5 0xf197509d in UTIL_SetModel(CBaseEntity*, char const*) () from /home/container/tf/bin/server_srv.so #6 0xf14969cf in CBaseAnimating::SetModel(char const*) () from /home/container/tf/bin/server_srv.so #7 0xf10b7e3c in CEconEntity::UpdateModelToClass() () from /home/container/tf/bin/server_srv.so #8 0xf10b7ec4 in CEconEntity::ReapplyProvision() () from /home/container/tf/bin/server_srv.so #9 0xf110b5da in CEconWearable::Equip(CBasePlayer*) () from /home/container/tf/bin/server_srv.so #10 0xf1237cfb in CTFWearable::Equip(CBasePlayer*) () from /home/container/tf/bin/server_srv.so #11 0xf190a142 in CTFPlayer::ManageRegularWeapons(TFPlayerClassData_t*) () from /home/container/tf/bin/server_srv.so #12 0xf190a468 in CTFPlayer::GiveDefaultItems() () from /home/container/tf/bin/server_srv.so #13 0xf190a643 in CTFPlayer::InitClass() () from /home/container/tf/bin/server_srv.so #14 0xf19062f3 in CTFPlayer::Spawn() () from /home/container/tf/bin/server_srv.so #15 0xf18ed56b in CTFPlayer::ForceRespawn() () from /home/container/tf/bin/server_srv.so #16 0xf18ecdf4 in CTFPlayer::HandleCommand_JoinClass(char const*, bool) () from /home/container/tf/bin/server_srv.so #17 0xf191cd9e in CTFPlayer::ClientCommand(CCommand const&) () from /home/container/tf/bin/server_srv.so #18 0xf1200cff in CTFGameRules::ClientCommand(CBaseEntity*, CCommand const&) () from /home/container/tf/bin/server_srv.so #19 0xf14e6217 in ClientCommand(CBasePlayer*, CCommand const&) () from /home/container/tf/bin/server_srv.so #20 0xebfec156 in __SourceHook_FHCls_IServerGameClientsClientCommand0::Func(edict_t*, CCommand const&) () from /home/container/tf/addons/metamod/bin/metamod.2.tf2.so #21 0xffa9d7f8 in ?? ()
In the console.log file which can be generated by adding -condebug to the command line, we witnessed these lines when it crashed:
Warning: Table DynamicModels is full, can't add models/workshop/player/items/all_class/hwn2016_toadstool_topper/hwn2016_toadstool_topper_Medic.mdl 590/ - tf_wearable: UTIL_SetModel: not precached: models/workshop/player/items/all_class/hwn2016_toadstool_topper/hwn2016_toadstool_topper_Medic.mdl L 05/02/2019 - 04:50:22: Engine error: 590/ - tf_wearable: UTIL_SetModel: not precached: models/workshop/player/items/all_class/hwn2016_toadstool_topper/hwn2016_toadstool_topper_Medic.mdl
Both servers were mostly vanilla and I found it strange that it was referring to a workshop model. We didn't have any custom models on the server. Therefore, I'd assume it was adding client's custom workshop items to the Dynamic Models table to perhaps check it.
I found that setting the following ConVars resolved the issue on both servers:
sv_allowupload 0 sv_pure -1
I'm unsure what CVar specifically fixed this, but I would recommend setting both of these ConVars to the values above. If I had to guess, it was sv_pure checking client's custom files and adding too many custom models to the Dynamic Models table. Though, I am not certain.
Just thought I'd put this out there because I couldn't find anything online regarding this crash and was honestly stumped until I saw the console.log messages.
I hope this helps other server owners.
Thank you!
-
Hi @Prfct,
Would you be able to perform a trace route to the server's IP which is 216.52.148.47.
In Windows, if you open Command Prompt, please execute the following command(s):
Trace Route
tracert 216.52.148.47
Ping
ping 216.52.148.47
A trace route would be the most beneficial. You can PM the results to me if you don't feel comfortable posting it here. Please also provide your IPv4 address to me which can be found here. I will perform a trace route from the server to your IPv4 address to see what's going on. It's possible your IP was added to NFO's firewall and I will need to open a ticket asking about this.
Thanks.
-
I made a project regarding how I'd like future Members to be accepted:
Though, with the things I would like implemented, I feel it'd be easier/more efficient not checking for VAC bans. I do feel if they have recent GFL bans, they should be manually approved. But I don't mind regardless.
This would be better if somebody made an application with IPS 4 that achieves these things.
Thanks.
-
Hello,
Our Rust server used the same map seed when wiping this week (614). I've noticed people on the server wanting a different map. I don't want to wipe the server again this week for obvious reasons. But I'm willing to try another map seed with next week's wipe. Hopefully eventually we'll rotate around a different seed each wipe.
I wanted some suggestions for map seeds.
Thanks!
-
Nice 😄
It's a shame Orange X3 and Mario Kart didn't take off like the others have, but we can bring them back in the future. We're just trying to get our foot into the door with TF2 again. Once we have a community/player base, we'll definitely try to tackle old beloved game modes like Orange X3, Mario Kart, TF2Ware, and Dodgeball again!
Thanks!
-
Rust Vanilla was successfully moved to GS08.
Thank you!
-
GS09's maintenance has been completed. I had a bit of a scare earlier when the machine kept refusing to accept my SSH connection after the first reboot. It turns out it just took 30 - 40 minutes to fully reboot due to the system updates we've made.
With that said, Rust Vanilla was successfully moved to GS08 and I am in the process of switching Garry's Mod Sandbox to Murder now.
Edit
Garry's Mod Murder was successfully setup:
Thank you for your patience everybody!
-
Starting maintenance on GS09 now.
-
Garry's Mod Purge has been moved to GS08. I will be moving Rust tonight since the machine will be going down for maintenance anyways.
Thank you.
-
Vultr is looking into why we aren't able to not announce to Telia via the Dallas PoP. They've confirmed our BGP communities are set correctly.
They are also looking into an issue with their hop with NTT causing high latency/ping which I am being routed to at the moment.
I will post back here once I have an update!
Thank you.
-
GS08 maintenance has been completed and servers are back online.
I plan to perform this maintenance on GS09 at some point tonight when the population has cooled down. This may be a good time to move servers as mentioned here.
Thank you. -
Hello everyone,
I just wanted to let everyone know we'll be performing emergency maintenance on GS08 and GS09 at some point today/tonight. We are currently already doing this for GS08.
Our GS09 machine started experiencing a kernel panic error which was reported here and won't go away until we restart the machine. We will be performing this maintenance to update our systems and make some modifications to our scripts to prevent this from happening again. Machine restarts will be required. Therefore, the servers will be shut down at some point.
@Dreae and I are confident the changes we will be putting in-place will resolve the issue we've run into.
Thank you for understanding.
-
- Popular Post
- Popular Post
Hello everyone,
I just wanted to quickly announce two servers moves we will likely be performing in the next week from GS09 to GS08. Today, we reached over 560 total players (peak) on all of our servers via the server list. We haven't seen this high of a number since Valve's servers broke in CS:GO resulting in a big spike of population on our CS:GO servers in January of 2017. I don't think we've seen this high of a number since early - mid 2016 and we are coming closer to reaching our all-time peak which is 1299 in 2015
(Still long ways to go, though).

I think we had a higher number at some point.
Anyways, with the burst of population today, our GS09 machine was overloaded. Therefore, I plan to move the following servers to our GS08 machine which has more powerful hardware (suited to handle higher-slot servers):
- Rust Vanilla - 92.119.148.13:28015
- Garry's Mod Purge - 92.119.148.7:27015.
Rust is at 60+ players at the moment while Purge is at ~50. I believe it'd benefit both servers moving to the new machine. The new machine's specs are:
- Intel i7-7700K @ 4.2 GHz.
- 64 GBs of RAM
- 500 GBs SSD
- 1 Gbps port.
- 15 TBs bandwidth.
- $120.00/m
I'm not sure when I will move Purge, but it will be when it's empty. I will most likely move Rust during the next (full) wipe which will be next Thursday (5/2). Since it's Vanilla, we can just copy the server CFG over and edit our PoP servers to direct the traffic to the new machine for that IP via Compressor. It should be really simple
I just wanted to let everyone know about this. I will post an announcement along with updates as I'm moving these servers.
Thank you!
-
On 4/24/2019 at 12:16 AM, Jerry Hat Trick said:
Things have gotten wonky since TF2 was introduced. Lots of lag on both servers, and TTT minecraft 24/7, which is VERY odd.
Not sure if its pertinent or not, but here are my latest results:
Bear in mind, I am near Chicago.
1 <1 ms 1 ms <1 ms x.x.x.x
2 1 ms 1 ms 1 ms x.x.x.x
3 24 ms 26 ms 23 ms xxxxxxxxx.net [x.x.x.x]
4 24 ms 24 ms 23 ms iplx-agw1.inet.qwest.net [75.160.90.9]
5 30 ms 30 ms 27 ms cer-brdr-02.inet.qwest.net [67.14.8.73]
6 30 ms 28 ms 27 ms ae-10.r07.chcgil09.us.bb.gin.ntt.net [129.250.8.173]
7 28 ms 28 ms 29 ms ae-7.r20.chcgil09.us.bb.gin.ntt.net [129.250.6.62]
8 49 ms 46 ms 45 ms ae-4.r23.dllstx09.us.bb.gin.ntt.net [129.250.4.152]
9 48 ms 46 ms 46 ms ae-6.r10.dllstx09.us.bb.gin.ntt.net [129.250.5.4]
10 46 ms 46 ms 45 ms ae-0.a01.dllstx09.us.bb.gin.ntt.net [129.250.3.244]
11 48 ms 46 ms 46 ms ae-0.choopa.dllstx09.us.bb.gin.ntt.net [128.241.219.54]
12 * * * Request timed out.
13 * * * Request timed out.
14 47 ms 47 ms 46 ms server5.us.gflclan.com [92.119.148.5]Trace complete.
C:\Users\Ron>
Your route and ping appear to be fine to the servers. When you're in-game, can you run net_graph 4 (type it in the console) and perhaps take a video or screenshot of it (video preferably so we can see what's going on as you play).
I've also un-hid your post as there was no reason to hide it. I've censored out the first three hops for you as well, so that information isn't out there (I doubt this is actually the reason why you hid your post since you hide every one of your posts, but yeah, just did it in-case).
Thanks.
-
Hello,
Due to this issue with Vultr and Telia (AS1299) along with seeing further evidence indicating that Telia is routing clients all over the place, we've decided to stop announcing to Telia altogether on the PoP servers that supported it. Unfortunately, we cannot stop announcing to Telia on our Dallas PoP still.
I have a ticket with Vultr right now asking why we can't stop announcing to Telia on our Dallas PoP.
I'm hoping this results in better routing. We're going to be looking into GTT (AS3257) and NTT (AS2914) to see if they'd suite as primary peers for our US PoP servers.
Thanks.
-
- Popular Post
- Popular Post
Division Leader:
- @Hamtaro (STEAM)
Server Manager:
- @Scott
Senior Admins:
- n/a
Server Admins
- @Gigi
- @AIexis
- @smithichie
- @StereoRaedioVA- @Crack Jacks
- @Brock_- @Alexis
Trial Admins:
- n/a
If you would like to Apply for Admin, please do so here.
We also ask that if you report a Player through Forum Messages, Please ONLY message the Server Manager or any Current Server Admins! -
Hi all,
I just wanted to make everyone aware of some of the routing issues that have been occurring recently, especially in Texas. As of right now, we primarily use Vultr to host our PoP servers (Point of Presence) on our new Anycast network. We can only use them at the moment since we don't have our own ASN right now. We need to use their public ASN and they're the only hosting provider we've found that have good prices with many locations around the world.
Anyways, one of Vultr's main peers is Telia (AS1299). Vultr peers directly with Telia which is demonstrated here. From what I've seen, Telia isn't a good peer. I've noticed Telia adding many unnecessary hops to routes in my opinion which increases the chance of packet loss and higher latency. There was even a Reddit post made about this regarding Black Ops 4 servers here.
Currently, my home network is going through Telia to get to our PoP servers. Recently, it started routing me to the Los Angeles PoP instead of the Dallas PoP (I am located in San Antonio, TX, so I should be routed directly to the Dallas PoP). With that said, beforehand, one of the last Telia hops was having high latency issues which is shown within my MTR here (hop #11 and ignore the annoying latency spikes, it's Spectrum and I don't see them with my UDP connections to our game servers).
All in all, it's safe to say that Telia doesn't appear to be a good peer. Now, I've seen some peers that Vultr does peer directly with that has better routes from what I've seen. For example, GTT (AS3257) and NTT (AS2914). However, no matter what I do with Vultr's BGP communities on our Dallas PoP (configured via BIRD), I cannot:
- Force our clients to go through GTT or NTT.
- Stop announcing to Telia altogether.
I believe we should be able to do this since Vultr appears to peer with them directly, at least according to the routing propagation graph linked above. Here is a list of Vultr's BGP communities showing that the BGP community '64600:XXXX' (X's referring to the ASN you want to not announce to). However, unfortunately, no matter what I try, I cannot change the routing. Perhaps it varies between each Vultr data center or depends on the client's ISP itself. I still have a lot to learn, haha
Anyways, I addressed the above issues to Vultr themselves and have been troubleshooting with them. Today I got a response in regards to myself being routed to LA instead of Dallas:
QuoteHello Christian,
Just so you know, other customers have indicated similarly situation regarding routing via Telia. Let me discuss it with the rest of the team and I will let you know first part of next week.
I'm hoping Vultr is able to do something about this, but I'm afraid Telia is an all-in-all bad peering provider right now and apparently has always been that way. With that said, I've noticed many other routing issues from other clients in the thread I made here. Vultr has very bad peering points, especially in Dallas and Seattle I've noticed.
The good news is, once we get our ASN we'll be able to choose other hosting providers that has better peering. I am investigating each hosting provider that suites our needs and looking at their peering. If the peers they mainly use has a good reputation, I'd say we should go with them. If we had the money (I'm really not sure how expensive this is and it's pretty complex), we could even go as far as getting our own IP transits. The difference between peering and IP Transits can be found here. However, I don't think this is something we can:
- Afford.
- Need.
Though, we may consider it if we ever need it I guess.
Overall, that's it. I just wanted to make everyone aware of the issue and I will continue to work to improve the network. However, we're kind of stuck and waiting on Vultr until we get our own ASN.
If you have any questions or even advice, please reply!
Thank you.
-
2 hours ago, Jerry Hat Trick said:
That specifically isn't normal and I plan to talk to @Dreae about it. It appears the server responds to any port on the IP via the A2S_INFO query. For example, if I add 92.119.148.3:27015 to HLSW (the server's actual port is 27015), things work fine. But I can also add 92.119.148.3:27017 and it'll show the same player count but the player list will be non-existent. I confirmed it doesn't duplicate within the Master Server though which is good.
Thanks.
-
I am also aware the Orange X3 GameTracker is currently under CS:S. This will be changed soon. We changed our CS:S Dust2 server into TF2 Orange X3 which is why it is currently showing under CS:S.
Thank you.
-
- Popular Post
- Popular Post
Hi everyone,
We are reintroducing our TF2 division. These servers are being hosted under our new network and ever since setting them up, have been doing fairly well. As of right now, there are no Division Leaders and Server Managers. The Directors and I will be managing the division for the time being until things stabilize and are proven to last.
Servers
2Fort
We've setup a 2Fort-only server with fast respawn two days ago. So far, it has reached 20+ players both days and filled up fairly quick. The server is basically vanilla apart from fast respawn and the 32-player limit.
Orange X3
GFL's classic Orange X3 server is back! The server is running the original Orange X3 map along with having fast respawn. Apart from the location, the server is just like how it was back in 2011 - 2013 when it was the top Orange X3 server in TF2 (got to 8th in the world for all TF2 servers at one point via GameTracker). This was one of the first successful servers within GFL and apart from the old TF2Ware server (which may come back, who knows?!), is the only server I've paid out of my pocket for back in 2011 (back when we had a private hosting provider on the West Coast of the US). I've honestly missed this server and I know many others have as well such as @Alejandro and @Ben. It got to 32/32 (full) the first night we've released it again (which was tonight). Therefore, I'm having confidence it'll do well.
Here's a video by myself showing how the server was back in 2012. The server is setup very similarly now in regards to general game play:
Mario Kart
Mario Kart is another addition we plan to add to our TF2 rotation if everything goes to plan. GFL had a very successful Mario Kart server setup back in 2011 - 2013. It contained things like jetpack which made the server enjoyable to many. I plan to set it up very similarly with the classic Mario Kart map and jetpack plugin (if there's an improved version of the old plugin, I will use that instead). This was personally one of my favorite servers to play on back in 2011 - 2013.
Here's a video I made back in 2012 showing how the server was:
Conclusion
We are starting out with these three servers to start. However, the first two have been off to a good start. I will continue to monitor their growth as time goes on and decide whether or not we want to open more servers. However, let's focus on these three first. I am also working to build an admin team for these servers as well and if the division seems like it'll be worth it in the long-run, we'll start looking for Division Leader(s) and Server Managers.
If you have any questions, please feel free to reply.
Thank you!
-
-
Hi all,
Our hosting provider we use with our new Anycast network (Nexril) is having maintenance this Saturday - Sunday:
QuoteJamesLast Saturday at 5:03 PM
@everyoneThis message is to notify you of an upcoming scheduled maintenance window taking place between April 20th, 2019 and April 21st, 2019. In light of our recent PDU failure, we will be standardizing and upgrading our power distribution system to ensure future failures are reduced and resolved as soon as possible. This will requiring powering servers off for a period of roughly 15-30 minutes. All steps will be taken to minimalize impact to customers' services.
Maintenance Window: Saturday, April 20th 10:00 PM - Sunday, April 21st 4:00 AM CST
Total Outage Time (Estimated): 15-30 Minutes
Affected Services: VPS (Power & Networking), Dedicated Server (Power & Networking), 1U/4U Colocation (Power & Networking), Half/Full-Rack Colocation (Network Only)If you have any questions regarding this message, please feel free to contact us by creating a ticket on our website.
Thank you.



Difficulties Connecting
in Zombie Escape
Posted
Another update, @Prfct can connect to the server via console. This is what I recommended:
Thanks.