All my files are updated to 1.3-1467, yet I'm still getting the message 'Web files are not matched with plugin version: All files need to be same version (1.3-1467) - try refreshing browser cache (shift-reload)' I'm not getting any errors in the console either.
I've been a long time dynmap user (I requested the topography renderer). But a recent switch from Bukkit to plain MCS has left me kind of reeling. On Bukkit I get about 40ms/tile for rendering. On plain I'm getting 300+ms/tile.
Dynmap 1.4 a2
MinecraftForge v6.6.0.497
MCS 1.4.7
Server is quad core 3.2ghz, 4GB of ram (3.5 to java).
Load average is 1.52 average, all from java. This is 0.1 average when dynmap is not crunching out maps. It actually appears like dynmap might be running the server out of memory. I can allocate more ram to test this theory. Up to perhaps 12GB. Currently it's pushed out into swap by 300MB.
I'm the only one on the server. Mods present are Buildcraft, RP2, and Dynmap, but with a non-modified map except the pre-fill. Map was pre-filled out to 2000 chunks in all directions. I asked dynmap to radius render 300 from the middle.
Speaking of, dynmap's helper text is not showing up on plain. Commands at the console, or in-game only return text when dynmap does something, not when you /dynmap help, or any other command. It just gives the status updates on chunks rendered.
Edit: MCS topped out at 4GB. I'll sort the 64 bit issue. But that doesn't solve the really long render times though.
I've been a long time dynmap user (I requested the topography renderer). But a recent switch from Bukkit to plain MCS has left me kind of reeling. On Bukkit I get about 40ms/tile for rendering. On plain I'm getting 300+ms/tile.
Dynmap 1.4 a2
MinecraftForge v6.6.0.497
MCS 1.4.7
Server is quad core 3.2ghz, 4GB of ram (3.5 to java).
Load average is 1.52 average, all from java. This is 0.1 average when dynmap is not crunching out maps. It actually appears like dynmap might be running the server out of memory. I can allocate more ram to test this theory. Up to perhaps 12GB. Currently it's pushed out into swap by 300MB.
I'm the only one on the server. Mods present are Buildcraft, RP2, and Dynmap, but with a non-modified map except the pre-fill. Map was pre-filled out to 2000 chunks in all directions. I asked dynmap to radius render 300 from the middle.
Speaking of, dynmap's helper text is not showing up on plain. Commands at the console, or in-game only return text when dynmap does something, not when you /dynmap help, or any other command. It just gives the status updates on chunks rendered.
Edit: MCS topped out at 4GB. I'll sort the 64 bit issue. But that doesn't solve the really long render times though.
I'm running this build on a Forge with all those mods (and about 10 more) on the same server with the same settings as my Bukkit server, and getting equivalent performance (within a few milliseconds per tile). What are the data you see from /dynmap stats?
I'm running this build on a Forge with all those mods (and about 10 more) on the same server with the same settings as my Bukkit server, and getting equivalent performance (within a few milliseconds per tile). What are the data you see from /dynmap stats?
well that's another 4+ hours of being unable to use the server due to needing to fullrender again, mike any way to tell which of these versions is expected to explode and which are worth trying? hope you get it fixed soon.
24 hours gone again, wasted in futility, welcome to minecraft modding, that is (unfortunately) our community.
(an ode to minecraft modding)
i keep making the same mistake, i think "oh, i've given them a few months, surely it'll be fixed by now" THAT is why i seem to ALWAYS be angry to some of you, my advice is to get the lead out.
there were nice things here, until the mod author threw a tantrum.
When I try to useversion 1.4-alpha-2 I always get a crash when a ChickenChunks chunk loader is in a world.
Alpha 1 works without problems.
The weird thing about the crash is that class l (a lowercase L) does exist in the server jar.
java.lang.NoClassDefFoundError: l
at a.<init>(CrashReport.java:27)
at a.a(CrashReport.java:319)
at yc.a(World.java:412)
at org.dynmap.forge.DynmapPlugin$WorldUpdateTracker.a(DynmapPlugin.java:1189)
at yc.i(World.java:735)
at codechicken.chunkloader.TileChunkLoaderBase.activate(TileChunkLoaderBase.java:126)
at codechicken.chunkloader.TileChunkLoaderBase.s(TileChunkLoaderBase.java:39)
at zz.a(Chunk.java:1051)
at zz.a(Chunk.java:1024)
at aam.a(AnvilChunkLoader.java:422)
at aam.a(AnvilChunkLoader.java:103)
at aam.a(AnvilChunkLoader.java:83)
at im.f(ChunkProviderServer.java:182)
at im.c(ChunkProviderServer.java:118)
at im.d(ChunkProviderServer.java:166)
at yc.e(World.java:524)
at yc.q(World.java:2790)
at codechicken.chunkloader.ChunkLoaderManager$ChunkLoaderOrganiser.revive(ChunkLoaderManager.java:361)
at codechicken.chunkloader.ChunkLoaderManager$ChunkLoaderOrganiser.revive(ChunkLoaderManager.java:330)
at codechicken.chunkloader.ChunkLoaderManager.revivePlayerLoaders(ChunkLoaderManager.java:797)
at codechicken.chunkloader.ChunkLoaderManager.tickEnd(ChunkLoaderManager.java:791)
at codechicken.chunkloader.ChunkLoaderEventHandler.tickEnd(ChunkLoaderEventHandler.java:39)
at cpw.mods.fml.common.SingleIntervalHandler.tickEnd(SingleIntervalHandler.java:22)
at cpw.mods.fml.common.FMLCommonHandler.tickEnd(FMLCommonHandler.java:138)
at cpw.mods.fml.common.FMLCommonHandler.onPostWorldTick(FMLCommonHandler.java:282)
at net.minecraft.server.MinecraftServer.r(MinecraftServer.java:688)
at ho.r(DedicatedServer.java:270)
at net.minecraft.server.MinecraftServer.q(MinecraftServer.java:598)
at net.minecraft.server.MinecraftServer.run(MinecraftServer.java:497)
at fy.run(SourceFile:849)
Finding a lot of hits on this that don't have to do with us - //www.google.com/search?client=ubuntu&channel=fs&q=%22[STDERR" target="" data-ensure-absolute>+java.lang.NoClassDefFoundError%3A+l%22&ie=utf-8&oe=utf-8"]https://www.google.com/search?client=ubuntu&channel=fs&q=%22[STDERR]+java.lang.NoClassDefFoundError%3A+l%22&ie=utf-8&oe=utf-8 - Not sure who to blame, at present, but the only way I see to get something like that would involve someone messing with class loaders and/or ASM type bytecode modding (as Forge itself does, as do most 'coremods'). Alpha 1 did not try to look up the block ID associated with the modified block, which is needed to avoid some blocks that like to say they're updated more or less constantly (active BC pipes and some other machines), which is where the class load exception is occuring (inside the vanilla/forge code when we ask for the block ID at the location that has been reported as being updated). I've got some ideas on how to possibly dodge this, although I've had no luck with specifically reproducing it (I've had to work with a couple of folks that have been able to just to try to figure out what little I know about it so far). The 'l' class is actually an obfuscated class that is part of the CrashReport handling (so, basically, it almost sounds like its crashing while trying to generate a crash report....).
well that's another 4+ hours of being unable to use the server due to needing to fullrender again, mike any way to tell which of these versions is expected to explode and which are worth trying? hope you get it fixed soon.
Not pointing fingers, but ChickenChunks has been the most consistent element in producing this (all cases I've seen to date that involve Forge trapping while we call it). I'll see if I can produce the problem with CC (I don't use it, although I do use CodeChickeenCore, NEI and WR-CBE). Are you guys using any specific features of CC (such as chunk loading blocks?)
Not pointing fingers, but ChickenChunks has been the most consistent element in producing this (all cases I've seen to date that involve Forge trapping while we call it). I'll see if I can produce the problem with CC (I don't use it, although I do use CodeChickeenCore, NEI and WR-CBE). Are you guys using any specific features of CC (such as chunk loading blocks?)
yeah actually we are making extensive use of spot loaders from chicken chunks.
Rollback Post to RevisionRollBack
24 hours gone again, wasted in futility, welcome to minecraft modding, that is (unfortunately) our community.
(an ode to minecraft modding)
i keep making the same mistake, i think "oh, i've given them a few months, surely it'll be fixed by now" THAT is why i seem to ALWAYS be angry to some of you, my advice is to get the lead out.
there were nice things here, until the mod author threw a tantrum.
I'd like to add to the chorus of people saying that 1.3 and 1.4 are incredibly taxing on the server. My server is running on a Phenom 2 x3 710, and we have serious performance issues with dynmap installed, which disappear when dynmap is removed.
I'd like to add to the chorus of people saying that 1.3 and 1.4 are incredibly taxing on the server. My server is running on a Phenom 2 x3 710, and we have serious performance issues with dynmap installed, which disappear when dynmap is removed.
1.4 Alphs 2 has done wonders for our server. I have it currently configured to emphasize server playing performance over rendering performance, and the playability has greatly improved. I would say play with your settings a bit, or if you want, I could post my config file to see if it helps you.
Thanks again Mike for the update render trigger options, and performance enhancements. Keep up the good work.
Folks,
I'll be releasing 1.4-alpha-3 sometime today - time permitting - which includes a VERY important fix for the chunk loading in the Forge versions. Specifically, the bug allowed the chunks-per-tick limit to be exceeded (significantly), which would definitely result in significant lag (and cause the maxchunkspertick setting to be relatively ineffective). It also includes a change that I believe will sidestep the exceptions some folks have reported with 1.4-alpha-2, which I believe are due to some sort of conflict with some other chunk loading/generating mods. At the very least, I'll get 1.4-alpha-3 out for Forge with 1.4.6/1.4.7, but I'd like to hit all the supported Forge versions (since the problem is common).
Hi, I'm using the version for bukkit. What is the most efficient way to reduce the plugin's cpu/mem usage?
Say the map was turned into a static image, with the only dynamic element being the player heads moving around. Would this decrease mem/cpu usage? If so, what are the config values i need to edit? Is disabling all the render triggers enough or do i need to do something else as well?
To update the render i'd just do /updaterender once a week or so - I don't need it constantly udpating.
Hi, I'm using the version for bukkit. What is the most efficient way to reduce the plugin's cpu/mem usage?
Say the map was turned into a static image, with the only dynamic element being the player heads moving around. Would this decrease mem/cpu usage? If so, what are the config values i need to edit? Is disabling all the render triggers enough or do i need to do something else as well?
To update the render i'd just do /updaterender once a week or so - I don't need it constantly udpating.
If you want to do that, here's the formula:
1) Turn off all the render-triggers (put '#' in front of all of them)
2) Plan on running fullrender - update is good so long as the areas being update are contiguous, but it will miss 'pockets' of updates when they are in the middle of areas that are otherwise not up to date.
3) Set fullrenderplayerlimit to 1 (or a low number) - this will cause the (weekly) fullrender to 'pause' when 1 or more players are logged in, and to resume when they leave**** Version 1.4-alpha-3 released ****
[Forge] Fix error with throttling of chunk loading which often causes the chunks-per-tick limit to be far exceeded (causing lag)
[Forge] Delay block ID lookup during 'blockupdate' trigger to avoid using Forge calls to get block IDs that appear to have issues when certain chunk loading/generating mods are present - avoid exceptions during WorldUpdateTracker processing
Improve debug information during calls to server thread (chunk loading, etc)
Improve parsing of Forge configuration files (prep for XyCraft support)
This release is quite recommended for anyone with lag complaints on Forge, as it will likely improve the situation significantly.
That sounds like you've got a firewall, and the port isn't opened (the timeout is the best clue - if there wasn't a firewall, and the port wasn't there, it'd fail immediately). Check the firewall settings on your server to be sure you've opened the port.
wait I'm confused... I defiantly opened the port, all my friends can join and play with me... are you saying I should make a separate port for Dynmap?
Wondering if there is a way to hide all players on the server, except for the player viewing the map?
and also, with this, is there a way to make it so you can see players within a certain block radius? So basically the player can only see him/herself on the map, unless they come within a certain radius of another player, then they would appear.
If not, these would be great features.
In either case, love the plugin. Thanks for your hard work.
(In regard to a mod that gives realistic animal genetics):
Would you really rather have bees that make diamonds and oil with magical genetic blocks?
... did I really ask that?
Wondering if there is a way to hide all players on the server, except for the player viewing the map?
and also, with this, is there a way to make it so you can see players within a certain block radius? So basically the player can only see him/herself on the map, unless they come within a certain radius of another player, then they would appear.
If not, these would be great features.
In either case, love the plugin. Thanks for your hard work.
If you enable logins, and set protected-player-info to true, folks using the web UI will only see their own positions (if they log in), and guests will see nobody. Exception is folks that have the dynmap.playermarkers.seeall permission (on bukkit), or operators, who will still see everyone. See https://github.com/webbukkit/dynmap/wiki/Web-ui-login-support-and-permissions
Dynmap 1.4 a2
MinecraftForge v6.6.0.497
MCS 1.4.7
Server is quad core 3.2ghz, 4GB of ram (3.5 to java).
Load average is 1.52 average, all from java. This is 0.1 average when dynmap is not crunching out maps. It actually appears like dynmap might be running the server out of memory. I can allocate more ram to test this theory. Up to perhaps 12GB. Currently it's pushed out into swap by 300MB.
I'm the only one on the server. Mods present are Buildcraft, RP2, and Dynmap, but with a non-modified map except the pre-fill. Map was pre-filled out to 2000 chunks in all directions. I asked dynmap to radius render 300 from the middle.
Speaking of, dynmap's helper text is not showing up on plain. Commands at the console, or in-game only return text when dynmap does something, not when you /dynmap help, or any other command. It just gives the status updates on chunks rendered.
Edit: MCS topped out at 4GB. I'll sort the 64 bit issue. But that doesn't solve the really long render times though.
I'm running this build on a Forge with all those mods (and about 10 more) on the same server with the same settings as my Bukkit server, and getting equivalent performance (within a few milliseconds per tile). What are the data you see from /dynmap stats?
MCS
2013-01-10 16:15:33 [INFO] [Minecraft] Tile Render Statistics:
2013-01-10 16:15:33 [INFO] [Minecraft] world.ct: processed=19, rendered=19, updated=8, transparent=0
2013-01-10 16:15:33 [INFO] [Minecraft] world.flat: processed=92, rendered=92, updated=8, transparent=0
2013-01-10 16:15:33 [INFO] [Minecraft] world.t: processed=1391, rendered=1391, updated=22, transparent=0
2013-01-10 16:15:33 [INFO] [Minecraft] TOTALS: processed=1502, rendered=1502, updated=38, transparent=0
2013-01-10 16:15:33 [INFO] [Minecraft] Triggered update queue size: 16
2013-01-10 16:15:33 [INFO] [Minecraft] Active render jobs: world
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk Loading Statistics:
2013-01-10 16:15:33 [INFO] [Minecraft] Cache hit rate: 83.10%
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk sets: created=1589, attempted=1589
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk: loaded=38579, attempted=64656
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk load times: 20439.55 msec (0.53 msec/chunk)
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk set load times: 175522.55 msec (110.46 msec/set)
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk set delay times: 155083.01 msec (97.60 msec/set)
2013-01-10 16:15:33 [INFO] [Minecraft] Chunk set exceptions: 0
2013-01-10 16:15:33 [INFO] [Minecraft] World tick list processing calls: 0
Running at 140msec. (Actually running good, but server stuttering pretty bad).
Bukkit
16:21:50 [INFO] Tile Render Statistics:
16:21:50 [INFO] world.ct: processed=673, rendered=673, updated=111, transparent=0
16:21:50 [INFO] world.flat: processed=1239, rendered=1239, updated=84, transparent=18
16:21:50 [INFO] world.t: processed=673, rendered=673, updated=114, transparent=0
16:21:50 [INFO] TOTALS: processed=2585, rendered=2585, updated=309, transparent=18
16:21:50 [INFO] Triggered update queue size: 20
16:21:50 [INFO] Active render jobs: world
16:21:50 [INFO] Chunk Loading Statistics:
16:21:50 [INFO] Cache hit rate: 36.37%
16:21:50 [INFO] Chunk sets: created=2167, attempted=2167
16:21:50 [INFO] Chunk: loaded=72346, attempted=93213
16:21:50 [INFO] Chunk load times: 66916.45 msec (0.92 msec/chunk)
16:21:50 [INFO] Chunk set load times: 119770.91 msec (55.27 msec/set)
16:21:50 [INFO] Chunk set delay times: 52854.46 msec (24.39 msec/set)
16:21:50 [INFO] Chunk set exceptions: 0
16:21:50 [INFO] World tick list processing calls: 0
76ms (No noticeable server lag).
Longer times for Bukkit, I have VM's running for work, maybe some disk play happening there.
http://pastebin.com/PDU5xcCH
well that's another 4+ hours of being unable to use the server due to needing to fullrender again, mike any way to tell which of these versions is expected to explode and which are worth trying? hope you get it fixed soon.
24 hours gone again, wasted in futility, welcome to minecraft modding, that is (unfortunately) our community.
(an ode to minecraft modding)
i keep making the same mistake, i think "oh, i've given them a few months, surely it'll be fixed by now" THAT is why i seem to ALWAYS be angry to some of you, my advice is to get the lead out.
there were nice things here, until the mod author threw a tantrum.
Finding a lot of hits on this that don't have to do with us - //www.google.com/search?client=ubuntu&channel=fs&q=%22[STDERR" target="" data-ensure-absolute>+java.lang.NoClassDefFoundError%3A+l%22&ie=utf-8&oe=utf-8"]https://www.google.com/search?client=ubuntu&channel=fs&q=%22[STDERR]+java.lang.NoClassDefFoundError%3A+l%22&ie=utf-8&oe=utf-8 - Not sure who to blame, at present, but the only way I see to get something like that would involve someone messing with class loaders and/or ASM type bytecode modding (as Forge itself does, as do most 'coremods'). Alpha 1 did not try to look up the block ID associated with the modified block, which is needed to avoid some blocks that like to say they're updated more or less constantly (active BC pipes and some other machines), which is where the class load exception is occuring (inside the vanilla/forge code when we ask for the block ID at the location that has been reported as being updated). I've got some ideas on how to possibly dodge this, although I've had no luck with specifically reproducing it (I've had to work with a couple of folks that have been able to just to try to figure out what little I know about it so far). The 'l' class is actually an obfuscated class that is part of the CrashReport handling (so, basically, it almost sounds like its crashing while trying to generate a crash report....).
Not pointing fingers, but ChickenChunks has been the most consistent element in producing this (all cases I've seen to date that involve Forge trapping while we call it). I'll see if I can produce the problem with CC (I don't use it, although I do use CodeChickeenCore, NEI and WR-CBE). Are you guys using any specific features of CC (such as chunk loading blocks?)
yeah actually we are making extensive use of spot loaders from chicken chunks.
24 hours gone again, wasted in futility, welcome to minecraft modding, that is (unfortunately) our community.
(an ode to minecraft modding)
i keep making the same mistake, i think "oh, i've given them a few months, surely it'll be fixed by now" THAT is why i seem to ALWAYS be angry to some of you, my advice is to get the lead out.
there were nice things here, until the mod author threw a tantrum.
Thanks again Mike for the update render trigger options, and performance enhancements. Keep up the good work.
I'll be releasing 1.4-alpha-3 sometime today - time permitting - which includes a VERY important fix for the chunk loading in the Forge versions. Specifically, the bug allowed the chunks-per-tick limit to be exceeded (significantly), which would definitely result in significant lag (and cause the maxchunkspertick setting to be relatively ineffective). It also includes a change that I believe will sidestep the exceptions some folks have reported with 1.4-alpha-2, which I believe are due to some sort of conflict with some other chunk loading/generating mods. At the very least, I'll get 1.4-alpha-3 out for Forge with 1.4.6/1.4.7, but I'd like to hit all the supported Forge versions (since the problem is common).
Say the map was turned into a static image, with the only dynamic element being the player heads moving around. Would this decrease mem/cpu usage? If so, what are the config values i need to edit? Is disabling all the render triggers enough or do i need to do something else as well?
To update the render i'd just do /updaterender once a week or so - I don't need it constantly udpating.
If you want to do that, here's the formula:
1) Turn off all the render-triggers (put '#' in front of all of them)
2) Plan on running fullrender - update is good so long as the areas being update are contiguous, but it will miss 'pockets' of updates when they are in the middle of areas that are otherwise not up to date.
3) Set fullrenderplayerlimit to 1 (or a low number) - this will cause the (weekly) fullrender to 'pause' when 1 or more players are logged in, and to resume when they leave**** Version 1.4-alpha-3 released ****
wait I'm confused... I defiantly opened the port, all my friends can join and play with me... are you saying I should make a separate port for Dynmap?
and also, with this, is there a way to make it so you can see players within a certain block radius? So basically the player can only see him/herself on the map, unless they come within a certain radius of another player, then they would appear.
If not, these would be great features.
In either case, love the plugin. Thanks for your hard work.
yes.
Dynamic maps uses a different port than the game does. So yes, you need to open ... was it 8123 by default? (Long since changed in my config)
* Promoting this week: Captive Minecraft 4, Winter Realm. Aka: Vertical Vanilla Viewing. Clicky!
* My channel with Mystcraft, and general Minecraft Let's Plays: http://www.youtube.com/user/Keybounce.
* See all my video series: http://www.minecraftforum.net/forums/minecraft-editions/minecraft-editions-show-your/2865421-keybounces-list-of-creation-threads
(In regard to a mod that gives realistic animal genetics):
Would you really rather have bees that make diamonds and oil with magical genetic blocks?
... did I really ask that?
No - you just need to do the same procedures you may have done to make your MC server port accessible to the port for Dynmap (8123, by default).
If you enable logins, and set protected-player-info to true, folks using the web UI will only see their own positions (if they log in), and guests will see nobody. Exception is folks that have the dynmap.playermarkers.seeall permission (on bukkit), or operators, who will still see everyone. See https://github.com/webbukkit/dynmap/wiki/Web-ui-login-support-and-permissions