Too many open files issue (jre/lib/ext/localedata.jar)

Adrian Ojeda 6 years ago

Hey there!

I'm having a weird issue on my traccar server. I have a client with almost 100 devices (tk103, gt06 and vt300). The platform works well but each 2 weeks the server hangs out and I have this error message on wrapper.log

INFO|1854/0|Service traccar|18-07-03 22:11:47|Uncaught exception by Thread[pool-6-thread-1,5,main]:
INFO|1854/0|Service traccar|18-07-03 22:11:47|java.lang.InternalError:java.io.FileNotFoundException: /usr/lib/jvm/java-1.8.0-openjdk-1.8.0.151-1.b12.35.amzn1.x86_64/jre/lib/ext/localedata.jar (Too many open files)"

If I restart the server I don't have issues until 2 weeks again.

I have optimized my OS following this link https://www.traccar.org/optimization/ but without luck, the same issue again.

I'm using a EC2 AWS instance (Amazon Linux AMI release 2017.03) m5.large (2 vCPU 8GB )

Thank you so much in advance!

Anton Tananaev6 years ago

Check open file limit for Traccar process on your system.

Anisha Vishnoi6 years ago

@Adrian Ojeda - Have you found the solution, I am facing the same issue ?

Adrian Ojeda 6 years ago

Hey @Anisha Vishnoi, unfortunately I couldn't fix it yet. The only solution I found so far is restart the server once a week :(

I tried incrementing the value of ulimit on the systems, but didn't work.

This is what I did:
open the file /etc/security/limits.conf
append this:
root soft  nofile 50000
root hard  nofile 50000

and to verify
ulimit -Hn
ulimit -Sn

Maybe can help you, please. let me know if you find something else!

Just as extra information I'm using a Ec2 instance in aws.

tecseguridad6 years ago

Hello i have same issue (Too many open files)

I tried incrementing the value of ulimit on the systems, but didn't work.

This is what I did:
open the file /etc/security/limits.conf
append this:
root soft nofile 50000
root hard nofile 50000

also try

  • soft nofile 50000
  • hard nofile 50000

and to verify
ulimit -Hn
ulimit -Sn

And always get for -Hn 4096 value and for -Sn 1024

Im already try several ways searching in formus, and did not work.

My System is CentOS Linux 7.6.1810

Can any one give me some light about this? Thank you

Adrian Ojeda 6 years ago

Well, I have followed this steps:
1 - Setup configuration files following this -> https://gist.github.com/luckydev/b2a6ebe793aeacf50ff15331fb3b519d
2 - run htop on the console to see the process running and get the traccar-server.jar process id
3- then run: cat /proc/<put the process id here>/limits
4- check the label "Max open files" if the limit is different than the one you specified on the step 1, try restarting traccar and go to step 2 again

Adrian Ojeda 6 years ago

One more thing, keep in mind you are changing the max files configuration for root user so when you are going to start/restart
the traccar server, be sure you are doing as root always with sudo e.g sudo /opt/traccar/bin/startDaemon.sh

tecseguridad6 years ago

Hi thanks for you answer, i try to open step 1 link but it doesn´t.

Can you please check if is ok?

Regards

Adrian Ojeda 6 years ago

weird.. let me copy here the content:

  • maximum capability of system
    user@ubuntu:~$ cat /proc/sys/fs/file-max
    708444

  • available limit
    user@ubuntu:~$ ulimit -n
    1024

  • To increase the available limit to say 200000
    user@ubuntu:~$ sudo vim /etc/sysctl.conf

  • add the following line to it
    fs.file-max = 200000

  • run this to refresh with new config
    user@ubuntu:~$ sudo sysctl -p

  • edit the following file
    user@ubuntu:~$ sudo vim /etc/security/limits.conf

  • add following lines to it

  • soft nofile 200000
  • hard nofile 200000
    www-data soft nofile 200000
    www-data hard nofile 200000
    root soft nofile 200000
    root hard nofile 200000
  • edit the following file
    user@ubuntu:~$ sudo vim /etc/pam.d/common-session

  • add this line to it
    session required pam_limits.so

  • logout and login and try the following command
    user@ubuntu:~$ ulimit -n
    200000

  • now you can increase no.of.connections per Nginx worker

  • in Nginx main config /etc/nginx/nginx.conf
    worker_connections 200000;
    worker_rlimit_nofile 200000;

Iván Ávalos5 years ago

I'm using an Amazon EC2 instance as well, with "Amazon Linux AMI release 2018.03" installed. I'm having the problem in just one of the two instances I have running. I don't really know what's the problem.