Mongo Restart Error -- /var/run/mongodb/mongod.pid exists - mongodb

Question

I just started a new AWS Linux AMI 2016.09.1(HVM) and downloaded the latest Mongodb release. I am using MongoDB version 3.0.14. I updated /etc/mongod.conf to the following,
# mongod.conf
# for documentation of all options, see:
# http://docs.mongodb.org/manual/reference/configuration-options/
# where to write logging data.
systemLog:
destination: file
logAppend: true
path: /var/log/mongodb/mongod.log
# Where and how to store data.
storage:
dbPath: /var/lib/mongo
journal:
enabled: true
# engine:
# mmapv1:
# wiredTiger:
# how the process runs
processManagement:
fork: true # fork and run in background
pidFilePath: /var/run/mongodb/mongod.pid # location of pidfile
# network interfaces
net:
port: 27017
# bindIp: 127.0.0.1 # Listen to local interface only, comment to listen on
all interfaces.
#security:
#operationProfiling:
#replication:
#sharding:
## Enterprise-Only Options
#auditLog:
#snmp:
I run Mongodb for the first time using service,
sudo service mongod start
Everything works fine. When I stop Mongodb and restart,
sudo service mongod stop
sudo service mongod restart
I get the following error,
Error starting mongod. /var/run/mongodb/mongod.pid exists.
Here is what is in the logfile, /var/log/mongodb/mongod.log,
2017-01-16T14:11:16.869+0000 I CONTROL ***** SERVER RESTARTED *****
2017-01-16T14:11:16.872+0000 I CONTROL [initandlisten] MongoDB starting :
pid=2820 port=27017 dbpath=/var/lib/mongo 64-bit hos
t=ip-10-0-0-91
2017-01-16T14:11:16.872+0000 I CONTROL [initandlisten] db version v3.0.14
2017-01-16T14:11:16.872+0000 I CONTROL [initandlisten] git version:
08352afcca24bfc145240a0fac9d28b978ab77f3
2017-01-16T14:11:16.872+0000 I CONTROL [initandlisten] build info: Linux
ip-10-63-190-181 3.4.43-43.43.amzn1.x86_64 #1 SMP Mon
May 6 18:04:41 UTC 2013 x86_64 BOOST_LIB_VERSION=1_49
2017-01-16T14:11:16.872+0000 I CONTROL [initandlisten] allocator: tcmalloc
2017-01-16T14:11:16.872+0000 I CONTROL [initandlisten] options: { config:
"/etc/mongod.conf", net: { port: 27017 }, processMan
agement: { fork: true, pidFilePath: "/var/run/mongodb/mongod.pid" },
storage: { dbPath: "/var/lib/mongo", journal: { enabled: t
rue } }, systemLog: { destination: "file", logAppend: true, path:
"/var/log/mongodb/mongod.log" } }
2017-01-16T14:11:16.893+0000 E NETWORK [initandlisten] Failed to unlink
socket file /tmp/mongodb-27017.sock errno:1 Operation
not permitted
2017-01-16T14:11:16.893+0000 I - [initandlisten] Fatal Assertion
28578
2017-01-16T14:11:16.893+0000 I - [initandlisten]
***aborting after fassert() failure
I have seen several blogs posted about this problem with solutions but none of them seem to work. Spent all day Sunday trying to figure this one out. Never had this problem before running an earlier AWS instance and Mongo.

Answer 1

Ok, found the problem. I deleted the /var/run/mongodb/mongod.pid file as suggested by Dave Maples above. However, MongoDB would still not run. I looked at the log file again and now saw it barking about this file,
/tmp/mongodb-27017.sock
It looked like it did not have the correct permissions. I tried,
sudo chown mongod:mongod /tmp/mongodb-27017.sock
And it worked fine. Now when I start/stop/restart MongoDB multiple times I no longer have the problem. I think the file permissions caused MongoDB to not clean itself up properly and left /var/run/mongodb/mongod.pid file in the directory. This caused the "Error starting mongod. /var/run/mongodb/mongod.pid exists" error. However, this was not the real problem. The real problem was the /tmp/ permissions. Looking at the log file after I deleted the mongod.pid file and restarting Mongod showed the permissions problem. So here is what fixed the problem,
sudo rm /var/run/mongodb/mongod.pid
sudo chown mongod:mongod /tmp/mongodb-27017.sock
I bet if I just tried "sudo chown mongod:mongod /tmp/mongodb-27017.sock" right after installing MongoDB on the new server before running MongoDB would have avoided the problem. Wasted a perfectly good Florida sunny day on this one.

Answer 2

Ok, found a better solution to what I had in my first answer. My previous answer simply eliminated the symptoms. I saw that several other people had to chown various other directories. I believe the problem is that mongod does not restart properly and leaves a number of directories incorrectly set up as it fails. I suspect multiple threads are running so what directory that gets incorrectly set up is somewhat random. I am using Amazon's AWS AMI Linux release and installing MongoDB 3.4. I found a very good MongoDB install script at
https://docs.mongodb.com/manual/tutorial/install-mongodb-on-amazon/
So here is the fix. I tested it several times and it worked each time. Install mongo as you normally would. Use the above script if helpful. Before you start mongod execute the following command,
sudo chkconfig mongod on
You will note that the above install script highlights chkconfig but at the bottom of the script where it might not get seen. This will ensure that you can successfully restart mongod each time. This should work on other linux versions.

Answer 3

you might change the ip in configuration file:
vim /etc/mongod.conf
# bindIp: 127.0.0.1 # Listen to local interface only, comment to listen on
all interfaces.
bindIp: xxx.xxx.xxx.xxx

Answer 4

You have another option in case of the assignation of user and group dosen't work but this depends directly of the state of sock file.
I moved the mongodb-27017.sock and renamed it, remove the .pid file and start the service and with this the process works
sudo mv /tmp/mongodb-27017.sock /tmp/mongodb-27017.sock2
sudo rm /var/run/mongodb/mongod.pid
sudo service mongod start

Answer 5

Try Rebooting the server , connecting to the server and run the command again. Worked fine for me.

Answer 6

After stopping the server, you are trying to restart. You can only start the server. That's why you are getting the stopping error. And also no need to worry about this error.