- Phase 1:
-
Python must be installed on the host machine to run this project
-
use the following command to run final-project, and then view the results in the
logs/final-project/
directory:bash run-ran.sh // read transaction files in random order // or bash run-rr.sh // read transaction files in round-robin order
-
-
Docker Engine must be installed on the host machine to run this project.
- if using linux, try the
install-docker.sh
script from the root directory:bash install-docker.sh
- if using linux, try the
-
The project is launched using
docker-compose
, which builds a container stack that includes a Postgres server and PgAdmin.-
Use the
run-seed.sh
script from the root directory to freshly build, launch, and seed the project (THIS WILL DELETE ANY PERSISTED DATABASE CHANGES MADE AFTER THE LAST SEED):bash run-seed.sh
-
Use the
run-build.sh
script from the root directory to persist the current database, but rebuild the db Docker images (use when db source code has been modified and needs updated, but not seeded):bash run-build.sh
-
-
After the containers have been spun up and initialized for the first time, the db will have been seeded with data.
-
There are queries available in the Postgres environment that can be run.
- In the following command, the .sql file name used under the
cs2550/db/query/
dir is{query-name}
- Use the
run-query.sh {query-name}
script from the root directory to run a query:bash run-query.sh test-query test-query2 ...
- In the following command, the .sql file name used under the