Gesendet: Mittwoch, 18. Mai 2016 um 10:43 Uhr
Von: "Martijn Brinkers" <martijn(a)ciphermail.com>
Betreff: Re: [Djigzo users] Problem after java-Update in CentOS 6
On 05/17/2016 06:16 PM, Martijn Brinkers wrote:
On 05/17/2016 06:02 PM, fatcharly(a)gmx.de wrote:
Hi, I`m using a ciphermail on a CentOS 6 fully
patched and it worked
fine for quite a while, but after an update to
java-1.6.0-openjdk-184.108.40.206-220.127.116.11.el6_7.x86_64 today, I found
this in the /var/log/djigzo.log: ...ffffffffff600000-ffffffffff601000
r-xp 00000000 00:00 0 [vsyscall] JVM process is
gone. JVM exited while loading the application. Launching a JVM...
JVM received a signal UNKNOWN (6). java version "1.6.0_39" OpenJDK
Runtime Environment (IcedTea6 1.13.11) (rhel-18.104.22.168.el6_7-x86_64)
OpenJDK 64-Bit Server VM (build 23.25-b01, mixed mode)
*** glibc detected ***
free or corruption (!prev): 0x0000000001e5ac c0 *** =======
Backtrace: ========= /lib64/libc.so.6(+0x75f4e)[0x7fc450e82f4e]
There were 10 failed launches in a row, each lasting less than 300
seconds. Giving up. There may be a configuration problem: please
check the logs. <-- Wrapper Stopped
On the Weblogin-page I got the message about a not fully started
Any suggestions are welcome
This looks like a bug in Java 6. A similar bug was reported last week:
For now the best solution is to upgrade to Java 6. See bug report
#GATEWAY-101 for more information.
A possible workaround is to add the following line to the file
Let me know whether this work around works for you. It looks like some
interaction between Java wrapper (which starts the Java process) and
Java 6 update.
I will investigate this further.
As a workaround (instead of installing Java 7), you can add
wrapper.disable_console_input=FALSE to the Java wrapper config. You can
add this option by editing the file /etc/init.d/djigzo
I'll contact the java wrapper developers. It might be some interaction
bug between changes in Java 6 and java wrapper.