The main goal is to get Pacha running and having you backing up your configuration files in as little steps as possible. More advanced Pacha functions are not covered here.
We recommend installing from the Python Package Index (PYPI):
pip install pacha
Have your SSH keys ready for the machines you want Pacha to talk to.
Even if you plan to keep files in a single server, get your ssh keys for localhost. Pacha pushes information only via SSH.
If this is a single machine, below is a quick example of getting SSH keys running:
cd ~/.ssh ssh-keygen [...] cat id_rsa.pub >> authorized_keys
Verify it is working by logging in passwordless:
After installing you need to add a config file. Pacha needs to have a configuration file added to be running properly. The config file can be located anywhere and can be called whatever you want. For a sample config file see Configuration
If no configuration file is added, Pacha complains:
+----------------------------------------------------+ | ** WARNING ** | | | | You have not set a configuration file for Pacha. | | To add a configuration file, run: | | | | pacha --add-config /path/to/config | | | +----------------------------------------------------+
Once you have a config file, run:
pacha --add-config /path/to/config
If you want to remove it, you can re-add it. Pacha will override the information.
And if you want to check the values that are being parsed, you can run:
pacha --config-values Configuration file: /Users/alfredo/vpacha/foo/pacha.conf log_level = DEBUG ssh_port = 22 hosts_path = /tmp/pacha/hosts host = localhost frequency = 60 master = True log_datefmt = %H:%M:%S ssh_user = alfredo log_format = %(asctime)s %(levelname)s %(name)s %(message)s
Mercurial needs to know what is the user that is going to make the changes. If you do not supply one the daemon will dye and you will get this error message next time you try to start:
Pacha searched for a Mercurial username in the $HOME directory and /etc/mercurial/hgrc but could not find one. Mercurial needs a username provided: But no username was supplied (see "hg help config") [ui] username = Firstname Lastname <email@example.com> verbose = True
So follow that above warning to add a username and avoid the daemon problem.
Pacha can run as a single instance but we want to accomplish a good configuration backup setup. Master/Slave is what we are going to cover here.
|host||Where is the Pacha master server running?. An IP or a FQDN works|
|port||If you are running a different port other than 22 for SSH, edit this. No need to specify if you have not changed the standard port.|
|ssh_user||The user Pacha will use to connect via SSH|
We will not cover the rebuilding process here. Again, the goal is to have Pacha backing up configuration files in this guide.
First we need to create a directory where all the configuration files will be pushed. This is the way of “granting permissions” in the Pacha server. So in the master server run:
pacha --add-host my_hostname
Replace my_hostname with the name of the machine you want to get configuration files from.
Pacha will not be able to push files if this is not done!
Pacha uses the –watch option to start tracking a directory. You can either be in the directory and run –watch or specify the path directly:
pacha --watch ~/bar adding foo.txt foo.txt running ssh alfredo@localhost "/usr/local/bin/hg init /tmp/pacha/hosts/mbp.local/bar" running ssh alfredo@localhost "/usr/local/bin/hg -R /tmp/pacha/hosts/mbp.local/bar serve --stdio" searching for changes 1 changesets found remote: adding changesets remote: adding manifests remote: adding file changes remote: added 1 changesets with 1 changes to 1 files
Although you have configured Pacha and added some files, the daemon process is not running. The daemon will help with the automated part of using Pacha.
You can start the daemon either in the background or foreground:
Background method (detaches from the terminal):
And in the foreground:
If you want to have more control and choices over the daemon process (like using Pacha with cron) check the Daemonization docs.