InsaneJ

Actually, it's my provider. They have migrated my connection to their new hosting platform last night. Unfortunately this also means I got a new IP address. I've changed all the DNS entries but it may take up to 24 hours before everybody can access the website and servers again.

The issue seems to have been resolved somehow. However I think the suspected player.dat file was included in today's backup. You may download it here: https://mega.co.nz/#!5kAnlahK!r2upqkjAIHJ0Pfo0pM4xhZc35Bc6sP85nh_LpZhTSbA

This server has been visited by 6118 individual players which started 956 towns. Some are more successful then others obviously, but most of the players on the server are doing just fine. If you can't make it there it's your fault, not the server's or anyone else's. Complaining about it won't change that.

You are either new to TFC or you are trolling. The South portal leads to the pole region where it's very cold and you will obviously find very little vegetation. This region is often used to help preserve food. The North, East and West Portals lead to plains areas with plenty of resources. The food won't fly into your mouth because almost 6000 players have been there before you. If you have to walk further then you'd like, you can also TP to another town and travel from there. Either way, this is TFC and it's supposed to be hard to get food.

Sorry to hear about your experience. We get hundreds of players all asking for food when they first join. How do you think the regular players feel about that? They have put in many hours getting settled in. Some feel like sharing and have put up food chests in their outposts at spawn. But some players don't feel like sharing. There's really nothing anybody can do about that. To help out new players there are 4 portals at spawn that lead to places far away that still have plenty of resources relatively close by. There's also a seed dispenser at spawn that drops seeds at dawn. You can use Dynmap to locate unpopulated areas. Even though a large portion of the map has been pre-rendered, most of the land mass is still relatively untouched. You shouldn't have a hard time finding a spot where you can settle down. All this information and more was already available to you at spawn if you had taken the time to read the signs. I hope you can make it with these tools at your disposal

Updated the server to the latest version of MCPC+. It's been a while since we've updated this for various reasons. Here's the changelog for those interested: Build 1.6.4-1.965.21.89: jdroque: Removed legacy getBukkitToForge method. This build also does not exist. Build 88: jdroque: Fixed config derp. Build 87: jdroque: Another fix to prevent hashCode check with null id. Build 86: jdroque: Prevent hashCode check with FakePlayer GameProfile's that contain a null id. GameProfile javadocs specify either an ID or name can be null/empty yet it uses the ID in its hashCode method. To workaround this issue, I now search for a matching gameprofile in HashMap, without causing a hashCode check, and create a new FakePlayer if no match is found. Build 85: jdroque: Fixed derp in config. This build doesn't exist =) Build 84: jdroque: Added configurable TileEntity tick settings. Refactored configuration. With this new change, each TileEntity that attempts to tick will be added to cauldron.yml under world-settings. There are currently 2 options for each TileEntity. The first option is 'tick-no-players' - This prevents ticks on a TileEntity if there are no players in range excluding forced chunks with chunkloaders. The default value is false. Setting this to true will make it tick with no players nearby. This setting should be set to true for TileEntities that need to always tick such as ComputerCraft's computer. The second option is 'tick-interval' - This controls how often you want this specific TileEntity to tick. The default is 1 which means you want it to tick every tick. Changing this to a value of 20 would mean it would tick every 20 ticks. Note: this option will only appear when a TE ticks. If the TE has never ticked before, you will not see this show up in config. You can still add the option manually to override it when it does tick. The default config will apply to all worlds. If you want to override a specific TE in a world, you would add it to cauldron.yml like the example below : world-settings: worldname: tileentities: net-minecraft-tileentity-TileEntityChest: tick-no-players: false tick-interval: 1 * added TileEntityCache to improve performance overall with new config. - Whenever a TileEntity ticks, if it hasn't been seen before, a new TileEntityCache will be created with the data found in config. For each new tick, the te config data will be accessed from the cache to avoid performance issues. If you modify the config, you need to reload by running '/cauldron reload'. This will update the TileEntityCache so it works immediately. * added caching for entity immediate despawn checks. * Fixed many configuration bugs. * You can now modify anything in cauldron.yml then execute '/cauldron reload' to have it take effect in game. For other configuration options use the '/cauldron' command. * Added StructureStart NPE check to prevent crashes with broken structure components. * ChunkGC is disabled again by default. If you experience a high chunk load, enable chunk-gc in bukkit.yml. Note: as stated in previous commits, you may experience issues with mods with ChunkGC turned on. Build 83: jdroque: Added TileEntityCache to improve performance overall with new config. Whenever a TileEntity ticks, if it hasn't been seen before, a new TileEntityCache will be created with the data found in config. For each new tick, the te config data will be accessed from the cache to avoid performance issues. If you modify the config, you need to reload by running '/cauldron reload'. This will update the TileEntityCache so it works immediately. * added caching for entity immediate despawn checks. Build 82: jdroque: Fixed config NPE with DimensionManager. Fixes #1333, #1335 Build 81: jdroque: Final build. Fixes #1327 The battle for GC being on and off is finally over, GC stays off. Agarian Skies servers recommended to turn on. IC2, EnderIO, BOP servers are recommended it remains off. To alter setting, use bukkit.yml and change chunk-gc.enabled to false or true. This is the final 1.6.4 build. jdroque: Added configurable TileEntity tick settings. Refactored configuration. With this new change, each TileEntity that attempts to tick will be added to cauldron.yml under world-settings. There are currently 2 options for each TileEntity. The first option is 'tick-no-players' - This prevents ticks on a TileEntity if there are no players in range excluding forced chunks with chunkloaders. The default value is false. Setting this to true will make it tick with no players nearby. This setting should be set to true for TileEntities that need to always tick such as ComputerCraft's computer. The second option is 'tick-interval' - This controls how often you want this specific TileEntity to tick. The default is 1 which means you want it to tick every tick. Changing this to a value of 20 would mean it would tick every 20 ticks. The default config will apply to all worlds. If you want to override a specific TE in a world, you would add it to cauldron.yml like the example below : world-settings: worldname: tileentities: net-minecraft-tileentity-TileEntityChest: tick-no-players: false tick-interval: 1 * Fixed many configuration bugs. You can now modify anything in cauldron.yml then execute '/cauldron reload' to have it take effect in game. For other configuration options use the '/cauldron' command. * Added StructureStart NPE check to prevent crashes with broken structure components. * ChunkGC is disabled again by default. If you experience a high chunk load, enable chunk-gc in bukkit.yml. Note: as stated in previous commits, you may experience issues with mods with ChunkGC turned on. Build 80: jdroque: Fixed server ping stalling. Build 78: jdroque: Updated README. Cleaned up some derp code. I need to sleep more and not work on bugs I already fixed =) Build 77: jdroque: Updated README. Cleaned up some derp code. I need to sleep more and not work on bugs I already fixed =) Build 76: jdroque: Don't remove players if detected in wrong chunk location. * AgarianSkies servers can suffer a big TPS loss with GC not enabled, so I have decided to just keep it on like MCPC+ always had. If you experience issues, you can set chunk-gc.enabled to false or period-in-ticks to 0. Build 75: jdroque: Don't remove players if detected in wrong chunk location. * AgarianSkies servers can suffer a big TPS loss with GC not enabled, so I have decided to just keep it on like MCPC+ always had. If you experience issues, you can set chunk-gc.enabled to false or period-in-ticks to 0.

Some more numbers The database in which we log all block modifications, inventory interactions, chat and everything else players do on the server has around 74 million records in it. During peak hours the database handles up to 400 queries per second. The database is around 9GB in size and holds data for 1 month. That may seem like a lot of trouble just for a Minecraft server. But it helps us to track down griefers, thieves and abusive players, resolve conflicts and keep the server a nice place for everyone.

Try restarting your Minecraft client that usually helps. If the Mojang authentication servers aren't on the fritz again that is.

Just got another report of the same problem. This person is using an Intel graphics card. But apparently if he installs Optifine the problem no longer occurs. He also suspects that somehow fruit trees are involved. Here are his crash-reports: http://paste2.org/Af2Lhj6J http://paste2.org/1Ommtvyv

Congrats on your 100th post

Looking good tenterro. Also love the huge rail road network you're building :)The HappyDiggers TFC server has reached a new milestone. We currently have had more than 4000 unique players visit our server so hooray for that!

The problem has occurred a few times after the upgrades for me now. And I've also gotten a report from one of the players on our server who uses the Tekkit launcher. He posted the following crash-report: http://pastebin.com/Bk4mh7QU

As far as we can tell the chunk corruption problem seems to be solved

EnTro triggered a solution I hadn't thought of before. I used Dinner Bone's Coordinate Tools to figure out which region files contained the corrupted chunks. Link: http://dinnerbone.com/minecraft/tools/coordinates/ Then I just downloaded the two suspected region files and used MCEdit to remove the suspected chunks from them and marked them for repopulation. When I opened the region files in MCEdit I noticed that a lot of items weren't removed from the chunks I removed in-game. Things that looked like vegetation was floating in mid-air. After putting the region files back on the server I visited the area and new chunks were being generated. However the landscape did look like Swiss cheese: And it generated a few odd chunks that were filled with ash doors for some reason. Also note the black squares on the mini-map where chunks didn't generate. During generation of these new chunks the server crashed once with a java.lang.StackOverflowError on the ash doors. Here's the crash-report, it's a whopper: http://paste2.org/ftWFpmD0 I manually regenerated all those chunks in-game and as far as we can tell the problem with the server crashing is now gone. Thanks for all the suggestions and tips

The past few days we've encountered a problem with our server where players would enter a corrupted chunk and then cause the server to crash.In the past we would simply use world edit to delete the affected chunk, then regenerate it, and the problem would go away.However now for unknown reasons the regenerated chunks still cause the server to crash when a player enters them.Here's a crash-report: http://paste2.org/xVe20fJP This is what it looks like in the server.log when a player joins a corrupted chunk: http://paste2.org/W43CwBDW Can anyone tell if this is a TFC or an MCPC+ or another mod/plugin issue? The only thing that has changed on the server from before we had this problem until now is the MCPC+ build. I've already tried running an older build with which we didn't encounter this problem. That didn't seem to make any difference. The past 2 days we've hadd 46 crashes because of this. Any help would be appreciated because right now the only alternative I'm seeing is deleting the world and starting over. Meaning the work of 3773 players spanning 657 towns would go to waste.

We have found that any mechanism or plugin that tries to edit a player's inventory is buggy when running TFC on MCPC+ This includes the /give command and plugins like openinv, but also creative mode middle-mouse clicking on a block to select that block. It turns up on your hot bar but you can't use the block. The only reliable way of giving players items so far is to just drop the items at their feet so they pick them up.

In an attempt to battle against the corrupted chunks that would crash the server, the following towns are now floating above the void and thus removed: InfinityCraftPTDerptopiaNovanikeybaseTerraFirmaLabs Sorry for you loss guys :B But if you are homeless and looking for a town, there are still 663 left to choose from

Velander, you're looking for MCPC+

Djakuta, posting a crash-report and forge mod loader log will help others to help you You can copy/paste the contents of log files to http://paste2.org and then give the URL here in this thread.

DeerGtus, Pernix is looking into your problem. Today the database got corrupted. I had to manually remove 2 towns that no longer had a mayor or residents. Also the town of Castle_Virrant had to be removed since I wasn't able to find and manually enter it's spawn point. P1eGuY and TheeShadows you will have to re-create your town and reclaim it's plots. I tried finding the information of that town in yesterday's backup. But since those backups are 4.4GB plain-text files (after extraction), I didn't feel like going through a weeks worth of backups. I've got better things to do on my day off. Sorry for the inconvenience. The server should be running now.

High CPU load in itself doesn't matter. What matters is the in-game server tick time. Try installing Opis on your server and your client. Then in-game OPs can use the /opis command which will give them a window outside of Minecraft with a ton of useful performance related information. In the summary tab you can see a real-time display of the server's tick time. This should always be below 50ms. When it goes above 50ms it means the server can no longer maintain 20 ticks per second. Installing Opis on your server does not force clients connecting to your server to also have Opis. You said you have 20GB of RAM available, which suggests you're currently using 12GB of RAM. That also suggests you're not running many services on that particular server. What kind of CPU does the server have?

Yes the ban is permanent. But don't worry, the AssHat server is back. You can steal from people all you like there.

I never play single player The problem also occurred with just Forge+TFC installed. I have updated Java to 7u60 and also installed Nvidia driver 337.88. So far the problem hasn't occurred since.

WorldBorder is definitely the way to go to pre-generate a map for TFC. You can (temporarily) use MCPC+ to run both Forge and Bukkit on the same server and pre-generate the map. Generating the map with players on can be either slow, or laggy. But never fast and without lag. So depending on your needs you may consider closing the server for a day or so to pre-generate the map.

Some more stats. The server continues to grow. The past 30 days the server has send+received 6.5TB of data: Also we have passed the 3000 unique player mark a while ago. Right now the server has been visited by 3233 players. Onwards to 4K