Upon restarting my Mac I got the dreaded Postgres error:
psql: could not connect to server: No such file or directory
Is the server running locally and accepting
connections on Unix domain socket "/var/run/postgresql/.s.PGSQL.5432"?
The reason this happened is because my macbook froze completely due to an unrelated issue and I had to do a hard reboot using the power button. After rebooting I couldn't start Postgres because of this error.
WARNING: If you delete postmaster.pid without making sure there are really no postgres processes running you, could permanently corrupt your database. (PostgreSQL should delete it automatically if the postmaster has exited.).
SOLUTION: This fixed the issue--I deleted this file, and then everything worked!
/usr/local/var/postgres/postmaster.pid
--
and here is how I figured out why this needed to be deleted.
I used the following command to see if there were any PG processes running. for me there were none, I couldn't even start the PG server:
ps auxw | grep post
I searched for the file .s.PGSQL.5432 that was in the error message above. i used the following command:
sudo find / -name .s.PGSQL.5432 -ls
this didn't show anything after searching my whole computer so the file didn't exist, but obviously psql "wanted it to" or "thought it was there".
I took a look at my server logs and saw the following error:
cat /usr/local/var/postgres/server.log
at the end of the server log I see the following error:
FATAL: pre-existing shared memory block (key 5432001, ID 65538) is still in use
HINT: If you're sure there are no old server processes still running, remove the shared memory block or just delete the file "postmaster.pid".
Following the advice in the error message, I deleted the postmaster.pid file in the same directory as server.log. This resolved the issue and I was able to restart.
So, it seems that my macbook freezing and being hard-rebooted caused Postgres to think that it's processes were still running even after reboot. Deleting this file resolved. Lots of people have similar issues but most the answers had to do with file permissions, whereas in my case things were different.
None of the above worked for me. I had to reinstall Postgres the following way :
Uninstall postgresql with brew : brew uninstall postgresql
brew doctor (fix whatever is here)
brew cleanup
Remove all Postgres folders :
rm -r /usr/local/var/postgres
rm -r ~/Library/Application\ Support/Postgres
Reinstall postgresql with brew : brew install postgresql
Start server : brew services start postgresql
You should now have to create your databases... (createdb)
If you're on macOS and installed postgres via homebrew, try restarting it with
brew services restart postgresql
If you're on Ubuntu, you can restart it with either one of these commands
sudo service postgresql restart
sudo /etc/init.d/postgresql restart
Maybe this is unrelated but a similar error appears when you upgrade postgres to a major version using brew; using brew info postgresql found out this that helped:
To migrate existing data from a previous major version of PostgreSQL run:
brew postgresql-upgrade-database
Here is my way:
launchctl unload ~/Library/LaunchAgents/homebrew.mxcl.postgresql.plist
rm /usr/local/var/postgres/postmaster.pid
launchctl load ~/Library/LaunchAgents/homebrew.mxcl.postgresql.plist
The following commands helped me out. The issue was with the PostgreSQL data version. Once upgraded, it started working fine for me.
brew upgrade postgresql
brew postgresql-upgrade-database
brew services restart postgresql
if your postmaster.pid is gone and you can't restart or anything, do this:
pg_ctl -D /usr/local/var/postgres -l /usr/local/var/postgres/server.log start
as explained here initially
For me, the solution was simply restart my computer. I first tried restarting with Brew services and when that didn't work, restarting seemed like the next best option to try before looking into some of the more involved solutions. Everything worked as it should after.
Another class of reasons why this can happen is due to Postgres version updates.
You can confirm this is a problem by looking at the postgres logs:
tail -n 10000 /usr/local/var/log/postgres.log
and seeing entries like:
DETAIL: The data directory was initialized by PostgreSQL version 12, which is not compatible with this version 13.0.
In this case (assuming you are on Mac and using brew), just run:
brew postgresql-upgrade-database
(Oddly, it failed on run 1 and worked on run 2, so try it twice before giving up)
Hello world :)The best but strange way for me was to do next things.
1) Download postgres93.app or other version. Add this app into /Applications/ folder.
2) Add a row (command) into the file .bash_profile (which is in my home directory):
export PATH=/Applications/Postgres93.app/Contents/MacOS/bin/:$PATH
It's a PATH to psql from Postgres93.app. The row (command) runs every time console is started.
3) Launch Postgres93.app from /Applications/ folder. It starts a local server (port is "5432" and host is "localhost").
4) After all of this manipulations I was glad to run $ createuser -SRDP user_name and other commands and to see that it worked! Postgres93.app can be made to run every time your system starts.
5) Also if you wanna see your databases graphically you should install PG Commander.app. It's good way to see your postgres DB as pretty data-tables
Of, course, it's helpful only for local server. I will be glad if this instructions help others who has faced with this problem.
This problema has many sources, and thus many answers. I've experienced each one of them.
1) If you have a crash of some sort, removing the /usr/local/var/postgres/postmaster.pid file is probably required as postgres may not have handled it properly. But ensure that no process is running.
2) Craig Ringer has pointed out in other posts that Apple's bundling of postgreSQL leads to pg gem installation issues Setting the PATH environment variable is a solution.
3) Another solution, is to uninstall and reinstall the gem. A brew update may be necessary as well.
If you stumble upon this post, if you can pinpoint one of the sources, you'll save time...
I was facing a similar issue here I solved this issue as below.
Actually the postgres process is dead, to see the status of postgres run the following command
sudo /etc/init.d/postgres status
It will says the process is dead`just start the process
sudo /etc/init.d/postgres start
This happened to me after my Mac (High Sierra) froze and I had to manually restart it (press and hold the power button). All I had to do to fix it was do a clean restart.
I had the same issue.
Most of the times, the problem is the fact that there's a leftover file
/usr/local/var/postgres/postmaster.pid
which works for most people, but my case was different - I tried googling this issue for last 3 hours, uninstalled postresql on OSX through brew, purged the database, nothing worked.
Finally, I noticed that I had an issue with brew that whenever I tried to install anything, it popped:
Error: Permission denied # rb_sysopen - /private/tmp/github_api_....
or something like it at the end of an install.
I simply did sudo chmod -R 777 /private/tmp and it finally works!
I'm writing this down because this might be a solution for someone else
I faced the same problem for psql (PostgreSQL) 9.6.11.
what worked for me -
remove postmaster.pid -- rm /usr/local/var/postgresql#9.6/postmaster.pid
restart postgres -- brew services restart postgresql#9.6
If postmaster.pid doesn't exist or the above process doesn't work then run --
sudo chmod 700 /usr/local/var/postgresql#9.6
For those running into this issue on M1 macs, try deleting this file and then restarting the brew service:
rm /opt/homebrew/var/postgres/postmaster.pid
My problem ended up being that I was using Gas Mask (a hosts file manager for Mac), and I didn't have an entry for localhost in the hosts file I was using.
I added:
127.0.0.1 localhost
And that resolved my problem.
I'm not entirely sure why, but my Postgres installation got a little bit screwed and some files were deleted resulting in the error OP is showing.
Despite the fact that I am able to run commands like brew service retart postgres and see the proper messages, this error persisted.
I went through the postgres documentation and found that my file /usr/local/var/postgres was totally empty. So I ran the following:
initdb /usr/local/var/postgres
It seems some configurations took place with that command.
Then it asked me to run this:
postgres -D /usr/local/var/postgres
And that told me a postmaster.pid file already exists.
I just needed to know if brew would be able to pick up the configs I just ran, so I tested it out.
ls /usr/local/var/postgres
That showed me a postmaster.pid file. I then did brew services stop postgresql, and the postmaster.pid file disappeared. Then I did brew services start postgresql, and VIOLA, the file reappeared.
Then I went ahead and ran my app, which did in fact find the server, however my databases seem to be gone.
Although I know that they may not be gone at all - the new initialization I did may have created a new data_area, and the old one isn't being pointed to. I'd have to look at where that's at and point it back over or just create my databases again.
Hope this helps! Reading the postgres docs helped me a lot. I hate reading answers that are like "Paste this in it works!" because I don't know what the hell is happening and why.
I had the same issue and it was due to an incompatible version after upgrading from version 11 to 13.2
Checking error log at:
/usr/local/var/log/postgres.log
Showed me:
DETAIL: The data directory was initialized by PostgreSQL version 11, which is not compatible with this version 13.2.
To fix I ran:
brew postgresql-upgrade-database
And then started postresql with brew:
brew services start postgresql
The causes of this error are many so first locate your log file and check it for clues. It might be at /usr/local/var/log/postgres.log or /usr/local/var/postgres/server.log or possibly elsewhere. If you installed with Homebrew you can find the location in ~/Library/LaunchAgents/homebrew.mxcl.postgresql.plist.
I had this same concern when connecting trying to start a PostgreSQL database server on MacOS Monterey.
When I run the command below to restart the PostgreSQL database server:
brew services restart PostgreSQL
It restarts but when I try to check the status of the PostgreSQL database server using the command below, I get an error:
Name Status User File
mysql started promisepreston ~/Library/LaunchAgents/homebrew.mxcl.mysql.plist
nginx started promisepreston ~/Library/LaunchAgents/homebrew.mxcl.nginx.plist
postgresql error 256 root ~/Library/LaunchAgents/homebrew.mxcl.postgresql.plist
Here's what worked for me:
First, I checked the log file for the PostgreSQL database server to what was the cause of the error using the command below:
cat /usr/local/var/log/postgres.log
OR
nano /usr/local/var/log/postgres.log
The logs showed the following errors:
"root" execution of the PostgreSQL server is not permitted.
The server must be started under an unprivileged user ID to prevent
possible system security compromise. See the documentation for
more information on how to properly start the server.
2022-01-25 19:01:06.051 WAT [29053] FATAL: database files are incompatible with server
2022-01-25 19:01:06.051 WAT [29053] DETAIL: The data directory was initialized by PostgreSQL version 13, which is not compatible with this version 14.1.
For the root execution error I had to run the following command to fix file permissions that when messed us when I ran brew services with the sudo command prefix. Replace your-username with your MacOS username (in my case my username was promisepreston:
# Stop postgresql
sudo brew services stop PostgreSQL
# In case service file is copied to ~/Library/LaunchAgents as well
brew services stop postgresql
# Fix permission of homebrew files
sudo chown -R your-username:admin $(brew --prefix)/*
For the database files are incompatible with server I had to simply upgrade the existing PostgreSQL data files which were created using version 13 to the latest PostgreSQL version on my computer which was 14.1 by running the following command below:
brew postgresql-upgrade-database
Afterwhich, I restarted the PostgreSQL database server:
brew services restart PostgreSQL
And then checked the status using the command below:
brew services list
Then I got the output below showing that everything was working fine:
Name Status User File
mysql started promisepreston ~/Library/LaunchAgents/homebrew.mxcl.mysql.plist
nginx started promisepreston ~/Library/LaunchAgents/homebrew.mxcl.nginx.plist
postgresql started promisepreston ~/Library/LaunchAgents/homebrew.mxcl.postgresql.plist
References: Brew PostgreSQL Starts But Process Is Not Running
I had a similar error.
All of this is done from the command line (no sudo calls at all)
I verified I had PostgreSQL installed psql -V (note that's a capital "V")
I attempted to connect to the server: psql postgres
THIS IS WHERE I EXPERIENCED THE ERROR OF THIS STACK OVERFLOW QUESTION
After doing some research about possible fixes, I obviously had PostgreSQL installed, but I didn't have a default server in place.
What I had to do was create a Custom Data Directory
As far as I can tell, creating the custom data directory is the same as having a default server in place.
Since this is a new machine (MacBook Pro 2021 using apple m1 chip), I wanted to find the easiest solution possible, and I believe this Custom Data Directory is just that. The remaining steps to fix this issue are as follows:
From the home directory, I created an empty directory mkdir myData
From the home directory, Initialized a server: initdb myData (throws a bunch of files into the myData directory)
pg_ctl -D myData -l logfile start (starts the server)
psql postgres (connects to the server)
So, as someone fairly new to PostgreSQL and databases and SQL in general, couple notes:
It is possible to "quit" the connection to the server, using \q (while connected to the server, it's also possible to type "help")
It is also possible to "stop" the server, as well with pg_ctl -D myData stop
At this point I now am certain I have PostgreSQL installed, have a server I can start and stop, and have the ability to connect to/disconnect from, that server.
Go to /var/log/
and run cat postgres.log
Here you will find the reason for the failure of postgres.
If it is a smart shut down then probably your icu4c version (C++ library for Unicode) is not proper which is linked with postgres. So run the following commands.
brew upgrade
brew cleanup
This should work ;)
I've had to look up this post several times to solve this issue. There's another fix, which will also applies to similar issues with other running programs.
I just discovered you can use the following two commands.
$top
This will show all the currently running actions with their pid numbers. When you find postgres and the associated pid
$kill pid_number
I just got the same issue as I have put my machine(ubuntu) for update and got below error:
could not connect to server: No such file or directory Is the server
running locally and accepting connections on Unix domain socket
"/var/run/postgresql/.s.PGSQL.5432"?
After completing the updating process when I restart my system error gone. And its work like charm as before.. I guess this was happened as pg was updating and another process started.
SUPER NEWBIE ALERT: I'm just learning web development and the particular tutorial I was following mentioned I have to install Postgres but didn't actually mention I have to run it as well... Once I opened the Postgres application everything was fine and dandy.
#Jagdish Barabari's answer gave me the clue I needed to resolve this. Turns out there were two versions of postgresql installed while only one was running. Purging all postgresql files and reinstalling the latest version resolved this issue for me.
I removed /usr/lib from the LD_LIBRARY_PATH and it worked.
I was working in dockerfile postgres:alpine.
This answer worked for me: https://stackoverflow.com/a/45454567/15067545 on my ubuntu system.
Command: sudo service postgresql restart.
i'm running mongo 1.8.2 and trying to see how to cleanly shut it down on Mac.
on our ubuntu servers i can shutdown mongo cleanly from the mongo shell with:
> use admin
> db.shutdownServer()
but on my Mac, it does not kill the mongod process. the output shows that it 'should be' shutdown but when i ps -ef | grep mongo it shows me an active process. also, i can still open a mongo shell and query my dbs like it was never shutdown.
the output from my db.shutdownServer() locally is:
MongoDB shell version: 1.8.2
connecting to: test
> use admin
switched to db admin
> db.shutdownServer()
Tue Dec 13 11:44:21 DBClientCursor::init call() failed
Tue Dec 13 11:44:21 query failed : admin.$cmd { shutdown: 1.0 } to: 127.0.0.1
server should be down...
Tue Dec 13 11:44:21 trying reconnect to 127.0.0.1
Tue Dec 13 11:44:21 reconnect 127.0.0.1 failed couldn't connect to server 127.0.0.1
Tue Dec 13 11:44:21 Error: error doing query: unknown shell/collection.js:150
i know i can just kill the process but i'd like to do it more cleanly.
It's probably because launchctl is managing your mongod instance. If you want to start and shutdown mongod instance, unload that first:
launchctl unload -w ~/Library/LaunchAgents/org.mongodb.mongod.plist
Then start mongod manually:
mongod -f path/to/mongod.conf --fork
You can find your mongod.conf location from ~/Library/LaunchAgents/org.mongodb.mongod.plist.
After that, db.shutdownServer() would work just fine.
Added Feb 22 2014:
If you have mongodb installed via homebrew, homebrew actually has a handy brew services command. To show current running services:
brew services list
To start mongodb:
brew services start mongodb-community
To stop mongodb if it's already running:
brew services stop mongodb-community
Update*
As edufinn pointed out in the comment, brew services is now available as user-defined command and can be installed with following command: brew tap gapple/services.
If you installed mongodb with homebrew, there's an easier way:
List mongo job with launchctl:
launchctl list | grep mongo
Stop mongo job:
launchctl stop <job label>
(For me this is launchctl stop homebrew.mxcl.mongodb)
Start mongo job:
launchctl start <job label>
Simple way is to get the process id of mongodb and kill it.
Please note DO NOT USE kill -9 pid for this as it may cause damage to the database.
so,
1. get the pid of mongodb
$ pgrep mongo
you will get pid of mongo, Now
$ kill
You may use kill -15 as well
If you have installed mongodb community server via homebrew, then you can do:
brew services list
This will list the current services as below:
Name Status User Plist
mongodb-community started <your_user_name> /Users/<your_user_name>/Library/LaunchAgents/homebrew.mxcl.mongodb-community.plist
redis stopped
Then you can restart mongodb by first stopping and restart:
brew services stop mongodb
brew services start mongodb
I prefer to stop the MongoDB server using the port command itself.
sudo port unload mongodb
And to start it again.
sudo port load mongodb
Check out these docs:
http://www.mongodb.org/display/DOCS/Starting+and+Stopping+Mongo#StartingandStoppingMongo-SendingaUnixINTorTERMsignal
If you started it in a terminal you should be ok with a ctrl + 'c' -- this will do a clean shutdown.
However, if you are using launchctl there are specific instructions for that which will vary depending on how it was installed.
If you are using Homebrew it would be launchctl stop homebrew.mxcl.mongodb
The solutions provided by others used to work for me but is not working for me anymore, which is as below.
brew services stop mongodb
brew services start mongodb
brew services list gives
Name Status User Plist
mongodb-community started XXXXXXXXX /Users/XXXXXXXXX/Library/LaunchAgents/homebrew.mxcl.mongodb-community.plist
So I used mongodb-community instead of mongodb which worked for me
brew services stop mongodb-community
brew services start mongodb-community
This is an old question, but its one I found while searching as well.
If you installed with brew then the solution would actually be the this:
launchctl unload ~/Library/LaunchAgents/homebrew.mxcl.mongodb.plist
If the service is running via brew, you can stop it using the following command:
brew services stop mongodb
I have memcached installed via homebrew.
how do I start/stop the server?
Any command-line tools to interact with memcached?
does homebrew have a way of removing a package?
When you installed it, it put a file named homebrew.mxcl.memcached.plist in /usr/local/Cellar/memcached/$version/; you copy that file into ~/Library/LaunchAgents and then tell launchd to start it with launchctl load -w ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist.
If you were watching the console carefully during the brew install command, it should have said something about doing these exact steps. If you run brew info it'll re-print this information, e.g. brew info memcached.
You should probably read https://docs.brew.sh/Manpage -- it has more useful documentation including the brew remove command which will remove the package if you decide you don't want it.
Brew used to have a command brew services (https://thoughtbot.com/blog/starting-and-stopping-background-services-with-homebrew), now deprecated. Instead, to get always-correct advice, run this command:
brew info memcached
Via telnet: telnet localhost 11211
See also What are some useful tips/tools for monitoring/tuning memcached health?
brew remove memcached
Additionally you can run "brew info", if you have forgotten about the instructions.
→ brew info memcached
memcached 1.4.7
http://memcached.org/
Depends on: libevent
/usr/local/Cellar/memcached/1.4.6 (8 files, 156K)
You can enable memcached to automatically load on login with:
mkdir -p ~/Library/LaunchAgents
cp /usr/local/Cellar/memcached/1.4.7/com.danga.memcached.plist ~/Library/LaunchAgents/
launchctl load -w ~/Library/LaunchAgents/com.danga.memcached.plist
Or start it manually:
/usr/local/bin/memcached
Add "-d" to start it as a daemon.
http://github.com/mxcl/homebrew/commits/master/Library/Formula/memcached.rb
To restart:
If you have the memcached starting up with launchd and your plist file has
<key>KeepAlive</key>
<true/>
Then you can just kill the process and it will reboot automagically.
ps ux | grep memcached
pkill -f memcached
ps ux | grep memcached
To stop launchd from restarting automatically:
launchctl unload -w ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist
To add memcached to launchd again:
launchctl load -w ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist
To uninstall something installed by homebrew:
brew uninstall memcached
And you can also create aliases
alias memcached-start="launchctl load ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist"
alias memcached-stop="launchctl unload ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist"
alias memcached-restart="launchctl unload ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist;launchctl load ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist"
and after that you cant easy start/stop/restart memcached
memcached-start
memcached-stop
memcached-restart
You can also use Lunchy to set the start/stop. Lunchy is a wrapper written over launchctl. I’ve written a detailed post about this.
$ gem install lunchy
$ mkdir ~/Library/LaunchAgents
$ cp /usr/local/Cellar/memcached/$version/homebrew.mxcl.memcached.plist ~/Library/LaunchAgents/
To start memcached
$ lunchy start memcached
To stop memcahed
$ lunchy stop memcached
You can enable Memcached to automatically load on login.
This first line creates a symlink (symbolic link) from where Homebrew installed it to the LaunchAgents folder.
ln -sfv /usr/local/Cellar/memcached/1.4.17/homebrew.mxcl.memcached.plist ~/Library/LaunchAgents/
Then to launch it now:
launchctl load -w ~/Library/LaunchAgents/homebrew.mxcl.memcached.plist