This system contains an addon which allows to syncronize multiple cl-prevalence systems state.
This is useful when you have apps running on multiple machines and want them to have eventually consistent state.
However, there are two caveat:
- Consistency is evential.
- In some cases consistency could not be reached.
How does this work
First thing you need to provide for all app instances is a folder where they will store their prevalence databases and logs.
This folder can be inside of a Dropbox, for example.
Eventually, each app should see all other app's transaction logs.
When you call a
sync-with-other-masterson the prevalence system, it does few things:
- makes a snapshot of it database;
- searches not processed transaction logs from other masters;
- and executes transaction from each found log;
- removes himself logs, processed by all other masters.
sync-with-other-mastersshould be called not often then once per second, because otherwise filenames of transaction log backups will clash.
- Ensure this will work on the real world application.
2. Support old transaction logs deletion after they was processed by all apps. 2. Add a documentation. 3. ...
Provide instruction how to build or use your library.
How to build documentation
To build documentation, you need a Sphinx. It is documentaion building tool written in Python.
To install it, you need a virtualenv. Read this instructions how to install it.
Also, you'll need a cl-launch. It is used by documentation tool to run a script which extracts documentation strings from lisp systems.
Run these commands to build documentation:
virtualenv --python python2.7 env source env/bin/activate pip install -r docs/requirements.txt invoke build_docs
These commands will create a virtual environment and
install some python libraries there. Command
will build documentation and upload it to the GitHub, by replacing
the content of the