Frontline Users Migration Tool

FrontLine 1.12 introduces OIDC support.

Users wanting to migrate might have issues because LDAP usernames are case insensitive so we store them in minor case. On the contrary, OIDC usernames are case sensitive.

As a result: usernames might no longer match.

This tool help you rename users to match the proper case, or even change the username to a new unique mapped attribute needed with OIDC support enabled.

WARNING: This tool make a dump of the users table, please double check dump content before performing the migration. A backup of the database would be preferred (for single-node, by copying /var/lib/cassandra).

Prerequisites

• Access to the Cassandra cluster
• bash
• cqlsh and awk commands in path

How to use

FrontLine need to be stopped. Update environment.conf with your cassandra host and port. Optionally, you can specify username and password.

Retrieve $dir/data/username.csv file by running the following command in a terminal:

$ cd frontline-users-migration-tool
$ ./fetch-usernames.sh

A dump of the users table will be created under $dir/data/dump.csv, make sure it is valid. It must contains all users, one per row, following the format: username,email,firstname,lastname,hashed_password,role_by_team

(From LDAP users, the email, firstname, lastname and hashed_password fields will be empty)

Then, update usernames in $dir/data/usernames.csv to desired case-sensitive usernames. (one username per row)

When you're ready, start the migration by running the following command in a terminal:

$ ./migrate-usernames.sh

$dir/data/migrated.csv contains the new users table content, and your database has been updated.

Output example

Step 1

$ cd frontline-users-migration-tool
$ ./fetch-usernames.sh

Copying gatling.users table content to dump.csv
Using 11 child processes

Starting copy of gatling.users with columns [username, email, firstname, lastname, password, role_by_team].
Processed: 1 rows; Rate:       7 rows/s; Avg. rate:       7 rows/s
1 rows exported to 1 files in 0.154 seconds.

Extracting usernames from dump.csv to usernames.csv
Update usernames in $dir/data/usernames.csv, then run the migration command (migrate-usernames.sh).

Step 2

Update usernames in $dir/data/usernames.csv

Step 3

$ ./migrate-usernames.sh

Updating dump.csv with usernames.csv, result will be in migrated.csv
Checking duplicates in usernames.csv...
No duplicate found.
migrate-users-perform.sh: line 28: $dir/data/migrated.csv: Permission denied
Dump with updated usernames under migrated.csv
 does not exist.
Users table content is going to be dropped...

Continue? [y/N] 
$ y
Drop users table.

Inserting migrated users (in migrated.csv)
Using 11 child processes

Starting copy of gatling.users with columns [username, email, firstname, lastname, password, role_by_team].
Processed: 1 rows; Rate:       1 rows/s; Avg. rate:       2 rows/s
1 rows imported from 1 files in 0.441 seconds (0 skipped).

Errors codes

1 - Cassandra host or port not specified

2 - Duplicate username in $dir/data/usernames.csv