You are not a server admin error

[dsebastien]

Expected Behavior

Stable connection to CouchDB

Current Behavior

Sometimes, I receive the "You are not a server admin" error when performing requests using my back-end server. I suppose that it is due to the way that I create the DB connection.

I'm currently using the following code:

export const databaseConnection: ServerScope = nano({
  url: environment.databaseConfiguration.url,
  requestDefaults: {
    // enable cookie-based authentication
    // Reference: https://www.npmjs.com/package/nano#using-cookie-authentication
    jar: true,
    auth: {
      username: environment.databaseConfiguration.username,
      password: environment.databaseConfiguration.password,
    },
  },

Later, in my repository code, I simply use that constant to make database calls:

public ensureThatWorkspaceExists(workspaceId: string): Observable<boolean> {
    return from(this.databaseConnection.db.list()).pipe(
      switchMap((databasesList) => {
        const databaseExists = databasesList.filter((existingDatabase) => {
          return existingDatabase === workspaceId;
        });

        if (databaseExists.length === 0) {
          throw new GateTechnicalError(GateTechnicalErrorType.DATABASE_NOT_FOUND);
        }
        return of(true);
      }),
    );
  }

Context

I've tried to configure cookie authentication in order to streamline things, but I suppose that I'm not using it correctly?
What surprises me is that it seems to work, most of the time, except that:

• it seems to fail if my API server is started before CouchDB
• it seems to fail after a while (when the cookie expires??)

When it fails, I get the "You are not a server admin" response when I try to make calls. Restarting the back-end server (not Couch) fixes the issue temporarily.

What would you recommend to avoid having such issues?

Your Environment

• Version used: 8.2.2
• Browser Name and version: N/A
• Operating System and version (desktop or mobile): Linux (Ubuntu 20.20)
• Link to your project: N/A

[janl]

is this by chance against a cluster (and not a single node)?

[dsebastien]

No, it is a single node.
For the moment, I've disabled cookie authentication as the issue continued to occur from time to time.

[janl]

is this maybe related to your configured cookie session time?

[dsebastien]

I think so, yes. I actually disabled the following setting: "allow_persistent_cookies". I guess I wanted to be overprotective, by avoiding "forever" cookies, and instead created a situation where I have "never" cookies? ;-)