dondi / GRNsight

Web app and service for modeling and visualizing gene regulatory networks.
http://dondi.github.io/GRNsight
BSD 3-Clause "New" or "Revised" License
17 stars 8 forks source link

Onboard the team with a locally running database #988

Closed dondi closed 1 year ago

dondi commented 1 year ago

After discussing this issue during the meeting, the preferred approach for developers on the team will be to install and run their own copy of PostgreSQL on their respective devices

@Onariaginosa, @ahmad00m, and @Sarronnn will plan a meeting during the week to walk through process, with @dondi being available to consult just in case. Original data files have been linked to under separate cover and may need to updated to @Onariaginosa’s latest copy

We will also take advantage of this to review our database documentation and loader scripts in the database folder of the repository. The hope is to get the documentation to a point where new members can also walk through it in order to get up and running (perhaps something for @ntran18 to try in the future)

Onariaginosa commented 1 year ago

I updated the documentation for loading data into an existing database. it is currently located on my branch. I was thinking that maybe we should either create a wiki page on setting up a local psql database or add it to the almost empty README.md at the top level database folder.

Onariaginosa commented 1 year ago

We set up and loaded the data on everyone's machines. We found that the instructions are not very beginner friendly, so I intend to update the top level database folder's README.md file and give play by play instructions on how to set up a local database.

Onariaginosa commented 1 year ago

I updated the documentation of the top level database folder's README.md so it gives full and in depth instructions on how to set up a local database instance

dondi commented 1 year ago

The created documentation looks really good and precise; @ahmad00m will try this on a different computer to test out how it feels when starting fresh

Database documentation can be a slippery slope because it does require a certain level of familiarity to begin with; otherwise the amount of explanation/background needed could expand excessively. So one potential addition to the docs would be a “Prerequisites/Assumptions” section stating the expected familiarity that the reader might have. Some supplementary third-party links may also be useful in order to introduce basic concepts about databases that we prefer to assume is already known to the reader

The link-out to Linux instructions looks good, and again that is hard to cover fully anyway due to the wide range of Linux distributions out there

Minor note for the future: namespace choices can probably be revised to have a more permanent feel (i.e., remove semester timestamps) as this feature settles in

We’ll monitor the Intermine error to see if it was just a one-off or something that might be worth reporting to the package maintainers

dondi commented 1 year ago

Also it sounds like the team successfully completed installation, but live testing for successful interaction with the local database has not been verified; that would be part of this issue before we can close it out

dondi commented 1 year ago

A common issue reported by both @ahmad00m and @Sarronnn was that the local web app still appears to not connect successfully to the local database. @dondi pointed out that the server/config/config.js file might not be sufficiently updated yet for the development configuration. So the group should look at that next in order to accomplish the ultimate goal which is to be able to run the GRNsight software alongside a locally-installed database

dondi commented 1 year ago

We were able to resolve this issue; it turns out that the pg library needed to be updated on @dondi’s, @ahmad00m’s, and @Sarronnn’s devices. The database was successfully accessed after that

Committed version was 8.0.0 and updating to 8.8.0 resolved the local database problem

However, a follow-up question remains regarding why @Onariaginosa had an operational stack even while on pg 8.0.0. The two variables that were identified were:

Everyone had the same PostgreSQL version and the same versions of the other libraries. We will close this issue but open another one that will document this versioning question