Unable to find PID of Tomcat (RHEL) - linux

I am Using Linux rhel6 2.6.32-220.el6.x86_64 #1 SMP Wed Nov 9 08:03:13 EST 2011 x86_64 x86_64 x86_64 GNU/Linux.
I can access tomcat from browser at port 8080. I am trying get PID of this .
I tried all the folling command but unable to fing the PID
lsof -i :8080
no output
netstat -tulpn | grep :8080
no output
also there is no log nor any httpd process running.
Please help me the find this. Thanks in advance

The tomcat application uses Java so that's why there is no httpd process running:
According to the source posted below the PID is posted in a $CATALINA_PID file which means that all you need to do is read that file with (TomCat version 5 or higher)
cat $CATALINA_PID
And that should post the PID.
Source:
Is Tomcat running?

I use to find open ports programs like this:
# lsof -i -n -P | grep LIST | grep 8080
python 10335 xxxxxxxxx 3u IPv4 10663405 0t0 TCP *:8080 (LISTEN)

The PID is usually defined in the init script. So you should take a look at
/etc/init.d/tomcat
Many init scripts seem to use this command for $PID:
ps aux | grep org.apache.catalina.startup.Bootstrap | grep -v grep | awk '{ print $2 }'
echo $CATALINA_PID will only return the PID if $CATALINA_PID which is not necassary the case.
source for init

Related

Large number of open files Linux on Fluentd machine

I have several questions regarding a Ubuntu machine which I am working on.
First of all here is the output of max open files:
$ cat /proc/sys/fs/file-max
1024000
However when checking how many are currently open:
$ lsof | wc -l
2002129
How is this number so much higer than the max?
Second, I had a look at what these open files were. 99% of them are these enteries:
ruby2.7 2749468 2750970 worker-14 fluentd 62u IPv4 1356781066 0t0 TCP {URL REDACTED}:5044->ip-10-153-7-71.eu-central-1.compute.internal:13637 (CLOSE_WAIT)
There is a mix of CLOSE_WAIT and ESTABLISHED connections. But when I check with netstat or ss it does not show that many connections:
ss -a | wc -l
1447
What else can I do to investigate why fluentd is rejecting connections and the machine is killing connections which cause the log delivery to fail?
As a final example, this one is a little crude. Listing fd's in /proc doesnt show as many as how many lsof shows:
ls -l /proc/*/fd | wc -l
1109

netstat gives 1 as PID for all ssh connections

In my custom linux image using netstat 1.42 (2001-04-15) from the package net-tools 1.60 I am seeing different behaviors for different linux versions..
On Linux 5.4.28:
# netstat -tnp | grep ESTABLISHED
tcp 0 0 <local_ip>:22 <client_ip>:14321 ESTABLISHED 29342/1
tcp 0 40 <local_ip>:22 <client_ip>:22470 ESTABLISHED 28443/0
On Linux 5.10.65:
# netstat -tnp | grep ESTABLISHED
tcp6 0 52 <local_ip>:22 <client_ip>:22470 ESTABLISHED 1/init
tcp6 0 52 <local_ip>:22 <client_ip>:33679 ESTABLISHED 1/init
The thing to note here is that on Linux 5.10, netstat shows PID as 1 and process as init for all ssh sessions. But in older linux, I am seeing correct PIDs for all ssh sessions. I need to know the client_ip and the PID for every ssh session so I can kill the session for whichever client I want.
From online searches, I could not figure out any fixes.
One alternative that I can see till now is using who to get this output:
# who
root pts/1 00:00 Jan 12 07:16:19 <client_ip>
root pts/2 00:03 Jan 12 08:21:07 <client_ip>
grep this with client_ip then use ps and grep that with the second column, like ps | grep pts/1 for first client.
Is there anything I can do to netstat to fix this though? Because netstat also give the client port which is useful for me.

Finding out Docker Container Running

I have a Linux machine at the company I work, and we know some web services are running on port 8111. I need to find out what is running check why the code is failing.
I do a netstat -at and I can see someone listening on the port:
tcp6 0 0 [::]:8111 [::]:* LISTEN
then I do a sudo lsof -i:8111 and I get the proc that is doing the listening:
dockerd 64285 root 266u IPv6 197189396 0t0 TCP *:8111 (LISTEN)
It's the docker daemon, so I do a docker ps -a but I don't see in the list a container that is listening on that port.
The person who was responsible for this abandoned the company and left no documentation, and just quit from one day to the next, and we are unable to get in touch with him.
Update:
To get all Ids, Image names with Ports from running instances try:
docker ps | tail -n +1 | awk '{print $1}' | xargs -n1 docker inspect | grep -E '"Id"|"Image"|"Ports"'

Is tftpd-hpa not a process in Linux?

I installed tftpd-hpa and tftp in Ubuntu 12.04 LTS.
when I check the status of tftpd-hpa with service command, it shows running.
zlf#ubuntu:~$ service tftpd-hpa status
tftpd-hpa start/running
But when I check its process, there is no running PID to show.
zlf#ubuntu:~$ ps -ef |grep ftp
zlf 5607 5397 0 15:02 pts/1 00:00:00 grep --color=auto ftp
zlf#ubuntu:~$ ps -ef |grep hpa
zlf 5609 5397 0 15:02 pts/1 00:00:00 grep --color=auto hpa
what happened? Is tftpd-hpa not a process or other reason?
By the way, I test tftpd-hpa by localhost, it doesn't work.
zlf#ubuntu:~$ tftp localhost
tftp> get test
Transfer timed out.
TFTP requests are probably handled by inetd/xinetd.

How to kill a process running on particular port in Linux?

I tried to close the tomcat using ./shutdown.sh from tomcat /bin directory. But found that the server was not closed properly. And thus I was unable to restartMy tomcat is running on port 8080.
I want to kill the tomcat process running on 8080. I first want to have the list of processes running on a specific port (8080) in order to select which process to kill.
This fuser 8080/tcp will print you PID of process bound on that port.
And this fuser -k 8080/tcp will kill that process.
Works on Linux only. More universal is use of lsof -i4 (or 6 for IPv6).
To list any process listening to the port 8080:
lsof -i:8080
To kill any process listening to the port 8080:
kill $(lsof -t -i:8080)
or more violently:
kill -9 $(lsof -t -i:8080)
(-9 corresponds to the SIGKILL - terminate immediately/hard kill signal: see List of Kill Signals and What is the purpose of the -9 option in the kill command?. If no signal is specified to kill, the TERM signal a.k.a. -15 or soft kill is sent, which sometimes isn't enough to kill a process.).
Use the command
sudo netstat -plten |grep java
used grep java as tomcat uses java as their processes.
It will show the list of processes with port number and process id
tcp6 0 0 :::8080 :::* LISTEN
1000 30070621 16085/java
the number before /java is a process id. Now use kill command to kill the process
kill -9 16085
-9 implies the process will be killed forcefully.
Option 1 A One-liner to kill only LISTEN on specific port:
kill -9 $(lsof -t -i:3000 -sTCP:LISTEN)
Option 2 If you have npm installed you can also run
npx kill-port 3000
One liner, a time saver
kill -9 $(lsof -t -i tcp:8080)
Explanation here: use a combination of lsof and kill
root#localhost:~# lsof -i tcp:8080
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
java 23672 sine 238u IPv6 3028222 0t0 TCP localhost:http-alt (LISTEN)
select pid and use kill
kill 23672
You can use the lsof command.
Let port number like here is 8090
lsof -i:8090
This command returns a list of open processes on this port.
Something like…
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
ssh 75782 eoin 5u IPv6 0x01c1c234 0t0 TCP localhost:8090 (LISTEN)
To free the port, kill the process using it(the process id is 75782)…
kill -9 75782
This one worked for me.
here is the link from the original post: link
If you want to kill a process running on port number 8080 then first you need to find the 8080 port process identification number(PID) and then kill it. Run the following command to find 8080 port number PID:
sudo lsof -t -i:8080
Here,
sudo - command to ask admin privilege(user id and password).
lsof - list of files(Also used for to list related processes)
-t - show only process ID
-i - show only internet connections related process
:8080 - show only processes in this port number
So you can now easily kill your PID using following command:
sudo kill -9 <PID>
Here,
kill - command to kill the process
-9 - forcefully
You can use one command to to kill a process on a specific port using the following command:
sudo kill -9 $(sudo lsof -t -i:8080)
For more you can see the following link
How to kill a process on a specific port on linux
Best way to kill all processes on a specific port;
kill -9 $(sudo lsof -t -i:8080)
This prints to stdout the process ids of everything running on <port_number>:
fuser -n tcp <port_number>
It also prints some stuff to stderr, so:
fuser -n tcp <port_number> 2> /dev/null
We can then supply these process ids to the kill command:
sudo kill $(fuser -n tcp <port_number> 2> /dev/null)
You could also put this in a function if you do it a lot:
function killport() {
sudo kill $(fuser -n tcp $1 2> /dev/null)
}
To know the pid of service running on particular port :
netstat -tulnap | grep :*port_num*
you will get the description of that process. Now use kill or kill -9 pid. Easily killed.
e.g
netstat -ap | grep :8080
tcp6 0 0 :::8080 :::* LISTEN 1880/java
Now:
kill -9 1880
Remember to run all commands as root
Get the PID of the task and kill it.
lsof -ti:8080 | xargs kill
First you need to do is run (replace with your port number):
fuser -k 3000/tcp
This will release the port. After you run the above command run:
service docker restart
And your problem is resolved.
try like this,
sudo fuser -n tcp -k 8080
Choose the port number and apply the grep in netstat command as shown below
netstat -ap | grep :7070
Console Output
tcp 0 0 :::7070 :::* LISTEN 3332/java
Kill the service based on PID ( Process Identification Number )
kill -9 3332
This will kill programs running on port 80
sudo fuser -k 80/tcp
Run the following command to find 8080 port number PID:
sudo lsof -t -i:8080
You can now easily kill your PID using following command:
sudo kill -9
You can use one command to to kill a process on a specific port using the following command:
sudo kill -9 $(sudo lsof -t -i:8000)
lsof -i tcp:8000
This command lists the information about process running in port 8000
kill -9 [PID]
This command kills the process
Linux: First you can find PID of this command if you know the port :
netstat -tulpn
example:-
Local Address Foreign Address State PID/Program name
:::3000 :::* LISTEN 15986/node
You then take the kill process. run the following command:
kill -9 PID
Expample: -
kill -9 15986
Linux: You can use this command if you know the port :
netstat -plten | grep LISTEN | grep 8080
AIX:
netstat -Aan | grep LISTEN | grep 8080
You then take the first column (example: f100050000b05bb8) and run the following command:
rmsock f100050000b05bb8 tcpcb
kill process.
You can know list of all ports running in system along with its details (pid, address etc.) :
netstat -tulpn
You can know details of a particular port number by providing port number in following command :
sudo netstat -lutnp | grep -w '{port_number}'
ex: sudo netstat -lutnp | grep -w '8080'
Details will be provided like this :
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
if you want to kill a process using pid then : kill -9 {PID}
if you want to kill a process using port number : fuser -n tcp {port_number}
use sudo if you are not able to access any.
Simply run this command. Don't forget to replace portnumber, with your port ;)
kill -9 $(sudo lsof -t -i:portnumber)
kill -9 `fuser 8080/tcp|xargs -n 1`, this commands also kills the process that listens on port 8080 with TCP connection
sudo apt-get install psmisc (or sudo yum install psmisc)
sudo fuser 80/tcp
Result:
80/tcp: 1858 1867 1868 1869 1871
Kill process one by one
kill -9 1858
I'm working on a Yocto Linux system that has a limited set of available Linux tools. I wanted to kill the process that was using a particular port (1883).
First, to see what ports we are listening to I used the following command:
root#root:~# netstat -lt
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State
tcp 0 0 0.0.0.0:hostmon 0.0.0.0:* LISTEN
tcp 0 0 localhost.localdomain:domain 0.0.0.0:* LISTEN
tcp 0 0 0.0.0.0:9080 0.0.0.0:* LISTEN
tcp 0 0 0.0.0.0:1883 0.0.0.0:* LISTEN
tcp 0 0 :::hostmon :::* LISTEN
tcp 0 0 localhost:domain :::* LISTEN
tcp 0 0 :::ssh :::* LISTEN
tcp 0 0 :::1883 :::* LISTEN
Next, I found the name of the process using port 1883 in the following way:
root#root:~# fuser 1883/tcp
290
root#root:~# ps | grep 290
290 mosquitt 25508 S /usr/sbin/mosquitto -c /etc/mosquitto/mosquitto.conf
12141 root 8444 S grep 290
As we can see above, it's the program /usr/sbin/mosquitto that's using port 1883.
Lastly, I killed the process:
root#root:~# systemctl stop mosquitto
I used systemctl becuase in this case it was a systemd service.
This is the solution for Windows:
C:\Users\Niroshan>netstat -ano|findstr "PID :8080"
Proto Local Address Foreign Address State PID
TCP 0.0.0.0:8080 0.0.0.0:0 LISTENING 18264
taskkill /pid 18264 /f
first check the process netstat -lt
check process id fuser <port number>/tcp
kill the process running on the port kill <process id>
Other way with Git Bash:
stopProcessByPortNumber() {
port=":${1}"
portStrLine="$(netstat -ano | findstr LISTENING | findstr $port)"
processId="$(grep -oP '(\d+)(?!.*\d)' <<< $portStrLine)"
echo $processId
taskkill -PID $processId -F
}
In my case cent os has some issue in suggested answer. So I used following solution :
ss -tanp | grep 65432 | head -1 | grep -Po "(?<=pid=).*(?=,)" | xargs kill
to build on what #veer7 said:
if you want to know what was on the port, do this before you kill it.
$ sudo netstat -plten |grep java
tcp6 0 0 127.0.0.1:8005 :::* LISTEN 1000 906726 25296/java
tcp6 0 0 :::8009 :::* LISTEN 1000 907503 25296/java
tcp6 0 0 :::8080 :::* LISTEN 1000 907499 25296/java
$ ps 25296
PID TTY STAT TIME COMMAND
25296 ? Sl 0:16 /usr/lib/jvm/java-8-openjdk-amd64/bin/java -Dcatalina.base=/hom
Use 'ps' and the number of the process that netstat reported back
How to kill process if the service on the port is not responding
timeout 1 telnet localhost 8080 2>&1 | if grep -q 'Unable'; then sudo kill -9 $(sudo lsof -t -i:8080); fi
What it does
Start 'timeout' with one second, to make telnet exit
' 2>&1 | ' pipe error-message without exiting
If error-message contains 'Unable', then run the kode to kill process locking port 8080
This can ble placed in a file and run regulary from 'sudo crontab'

Resources