A few of you might know Spigot (it's Minecraft server software). I'm trying to run their BuildTools.jar to update the current build, but I'm getting some weird git errors.
Info on BuildTools.jar
System: Windows 10. Worth noting I also have a few Google wifi things connected. My laptop is on the base wifi though and switching networks also gives the same errors.
Errors:
Exception in thread "main" org.eclipse.jgit.api.errors.TransportException: https://hub.spigotmc.org/stash/scm/spigot/bukkit.git: 504 Gateway Time-out when running command java -jar BuildTools.jar
fetch-pack: unexpected disconnect while reading sideband packet when running clone manually in PowerShell: git clone https://hub.spigotmc.org/stash/scm/spigot/bukkit.git
I have tried:
Running both clone and java jar commands in a mix of PowerShell, git bash, and Linux subsystem for Windows
git config --global http.postBuffer 524288000"
Opening TCP 9418 on my router
Reinstalling Java and Git
Tried cloning on a different Windows computer where I use git often, same errors
Google says:
Clone with SSH instead (as far as I know I can't do this without the Spigot admins adding support for it on the repo)
What can I try next?
Works for me without any issues on the first try (and very fast!):
Did you make sure you downloaded the latest version of all tools you use? Specially for java it can quickly happen to use an outdated version. From their instructions at least Java 8 is required. I am currently running Java 10 (OpenJDK).
Their docs also state:
Please be aware that it is required that you have BuildTools #35 or later, older versions will not work
So you should make sure you also have the appropriate version of the tools (which I expect since 35 seems to be really old).
Usually 5XX errors indicate server side errors. Maybe they had some config or network issues on their end. Therefore it might work if you'll try it again after waiting a little while.
Maybe just some network node on the way to their servers broke. Apart from waiting and to try again you could try to setup a VPN to route the traffic throgh to avoid this failure (or turn it off in case you're using some!).
Maybe it's also worth to check your firewall (or proxy) settings.
PS: I was using git bash also on a Win 10 machine.
Related
We are porting OpenJDK to RISC-V. We're at the point that the interpreter builds. We need to debug it, using GDB. However, we haven't been able to find a working GDB that works with RISC-V QEMU. Any help would be much appreciated.
It depends on what you mean by "GDB that works with QEMU". There's three use cases here
Running system-mode QEMU, trying to debug a machine-mode or supervisor-mode program (like the Linux kernel, for example). For this, you should be able to use the upstream RISC-V GDB port attached to QEMU's built-in GDB stup to debug like normal.
Running user-mode QEMU, trying to debug a user-mode program. I've never done this before, but I think you can still use QEMU's built-in GDB stub and connect with the upstream GDB.
Running system-mode QEMU, trying to debug a user-mode program (like OpenJDK, for example). In this case QEMU doesn't actually factor in to the equation, as you're really just looking for a standard Linux GDB port. Both our Linux and GDB ports are currently buggy when it comes to native debugging, but you should be able to take the latest upstream versions of both and at least get some work done. There's lots of issues with drivers in our upstream Linux port, but if you're just using QEMU's virt machine then you should be good to go.
I'm not actually sure StackOverflow is the best place to ask this question, as the answer will be obsolete pretty much as soon as it's been asked.
Environment:
Headless Linux server
Red Hat Enterprise Linux Server release 6.7 (Santiago)
2.6.32-573.12.1.e16.x86_64
I have a java(7) program that I run from command line that spawns multiple threads and hits a oracle database simultaneously using ojdbc7.
Intermittently I see connection reset error:
Could not get JDBC Connection; nested exception is java.sql.SQLRecoverableException: IO Error: Connection reset
exactly as reported here and resolved here.
I tried the following to mitigate the issue:
Add these variations to my command line
-Djava.security.egd=file:///dev/urandom
-Djava.security.egd=file:/dev/../dev/urandom
-Djava.security.egd=file:/dev/./urandom
Tried adding to java.security file
securerandom.source=file:///dev/urandom
securerandom.source=file:/dev/../dev/urandom
securerandom.source=file:/dev/./urandom
Also tried ojdbc6
But I still see the issue. Which means that even with urandom in use, still there is not enough realtime entropy. Which is understandable, because when I run my java program, everything else is stopped (this is a application server with multiple running jvms, usually. )
I am wondering if there is anything I can do to 'cause' entropy on the server. I have limited access to this server so I am limited in what I can do. This java job is expected to run for a couple of hours, so I can not keep typing on the keyboard etc. May be run a simple program in the background that 'does' something?
Any ideas? I tried rngd command but apparently I do not have permission to use it. Any help greatly appreciated, been stuck with this issue for a while now.
Edit:
I tried to run another java program (using java.awt.Robot) that simulates Keyboard presses continuously during my original program run, but due to limited permissions on the server I wasn't able to get it working. But in general, would that be a good way to ensure some entropy?
I have used Java on Windows 7 (64bit) machines for quite some time and have never had problems with not being able to create a network connection. Now on the systems at the company I'm working for I need to set "-Djava.net.preferIPv4Stack=true" or it seems the java processes are not able to create a single connection.
The downside is that if I don't set "_JAVA_OPTIONS" I would have to configure tons of services to use this setting. However if I use it Java outputs this silly "Picked up: _JAVA_OPTIONS..." to stderr (Wonder which guy made that silly decision). This however makes my GWT compiles fail in IntelliJ.
What I would like to know ... this is the first time I'm having these problems and I guess they must somehow be related to the setup of the Operating System. What is probably causing these problems (As I mentioned ... I have about 4 other Systems with windows 7 and 64 bit java vms that don't have these problems).
Chris
Ok so we found out the reason.
This problem seemed to have occured on systems containing a special version of remote access software that our company used. This seems to have inserted some modified dlls into Windows network stack (We were told in order to auto-detect network connectivity). It seems this dll had some issues, causing all IPv6 traffic to be blocked. It also caused the affected Machines to have regular Bluescreens. Uninstalling that software got the system Bluescreen free and I no longer need the preferIPv4Stack setting.
This is just a theory ...
According to the Java documentation, if IPv6 is available on the operating system, the underlying native socket will be an IPv6 socket.
If the operating system doesn't have IPv6, or it has properly configured IPv6, that's what you want. But if the operating system has IPv6 that is not working properly, then I imagine that Java will attempt to use IPv6 to make connections and fail.
However if I use it Java outputs this silly "Picked up: _JAVA_OPTIONS..." to stderr.
According to this Q&A, there is no way to turn it off: Suppressing the "Picked up _JAVA_OPTIONS" message
I would suggest setting _JAVA_OPTIONS globally (if you must) and then unsetting it for the environment you launch your IDE from.
Just my contribution:
Disabling all the ipv6 on the network devices solved it for me.
I hope this helps
I assume the question is strongly related to this one but it is closed and the solution does not seem clear to me.
I am trying to use VisualVM to profile a console Scala application of mine. I do this by starting VisualVM, starting my Scala application, opening its node in VisualVM, going to the "Profiler" tab and clicking "CPU". The button goes grey for some time (while status never changes to anything from "profiling inactive"), then back active but no profiling data appears shown.
The application does some file reading, data processing (which takes about a minute), stdout data output and exits. I've also tried adding Thread.sleep(60000) to the beginning and to the end of the program to ensure enough time for VisualVM to capture it and do the job, nothing changed though.
Unlike to the related question I've linked to I don't launch my app from Eclipse or anything else - I use the following command line to launch it:
java -classpath myapp.jar:lib/* MyApp.Main
all the libraries (including scala-library.jar) are put in lib/. The app works as expected.
Updates:
I have tried YourKit Java Profiler 11.0.2 and it fails with AttachNotSupportedException.
I've managed with YourKit by attaching the agent manually.
It seems worth stressing that I neither run the application from an IDE (or a build tool) nor have I modified any JVM options but classpath. As far as I understand this problem appears to depend on a JVM version and user/permission issues and. The objective is to figure out the actual mechanism of the problem and the way to configure all the things to work.
Versions of the software used:
YourKit Java Profiler 11.0.2
VisualVM 1.3.4
SBT 0.12.3 (to build but not to run (I've tried both actually))
Scala 2.9.2
Oracle Java 1.7.0_04-b20
XUbuntu Linux 12.04 32-bit i386
Linux kernel 3.2.0-24-generic-pae
I have used VisualVM to profile a Scala application two days ago, so I believe it should work. Here are my suggestions:
Take out YourKit out of the equation and run things without it for a starter
Take out scala out of the equation, try profiling a Java hello world that does while (true) { Thread.sleep(10000); }
Get that to work then put back Scala, then put back YourKit.
Edit: based on your input, I suggest to explicitly set the remote JMX connection. You can start java with the following options:
-Dcom.sun.management.jmxremote=true
-Dcom.sun.management.jmxremote.port=20000
-Dcom.sun.management.jmxremote.ssl=false
-Dcom.sun.management.jmxremote.authenticate=false
Then in VisualVM, add a JMX connection to localhost:20000.
JavaFX 2 is not support Linux yet. Does this mean a client Linux machine (user machine) cannot run it or a server Linux machine (host machine) cannot run it or both?
*EDIT:*
JavaFX is for rich client. So the server will not run it, but store it and client will get it and run it, right?
JavaFx depends on hardware acceleration to run. This is currently not supported on linux (expected 2012). So if your server is executing JavaFx code, then it would not run on the server. If it is only delivering code to clients, like in an applet, then it would work.
It means you can't use the JavaFX libraries on a machine running Linux. Neither a client Linux machine, nor a server Linux machine can use them.
Response to Edit
You can still store and serve JavaFX code on Linux machines. You just can't execute the JavaFX code on a Linux machine.
Note that JavaFX is different from Java. You can still use Java on Linux machines.
I realize this question was posed a while back (11/2011), but I thought it might help to point out the related Open Source project from the OpenJDK community called OpenJFX. Feel free to help out with development, if you're qualified to do so as well. Given the secure nature of Linux, this project should prove a secure/safe way to utilize the "hardware acceleration to run"; though, it's still in the development stages.
Take a look at the project wiki to learn how to build OpenJFX on Linux.