🇬🇧 Successful CMS Migration to Drupal with a Junior Team Under Tight Constraints
Summary:
In this video, I share how I successfully migrated a full CMS into Drupal with a junior team under a tight budget and deadline, all while ensuring no downtime for active content editors. I created a proof of concept that exposed a single endpoint for the site structure, allowing us to process content efficiently using Drash commands instead of the more complex Drupal Migration API. I also empowered the junior developers by providing them with a UML class diagram, documentation, and live training sessions, which made the migration sustainable. The project was completed successfully, and the client was extremely happy, even leaving a positive recommendation on my LinkedIn. I encourage you to consider these strategies for your own projects to enhance team capability and project success.
Transcript:
0:00: Hey, this is how I immigrated a full CMS into Drupal with a junior team and a brutal deadline without sacrificing quality.
0:11: Hehehe, yes, the software architect was unavailable, the budget was tight and the team available were mostly junior PHP developers who were just, were just studying into Drupal.
0:24: Also, the containers were still actively using the site. So, downtime of breaking content was not an option. The real challenge, listen.
0:40: Okay, so the real challenge wasn’t only technical. It was how do you deliver a safe migration fast? And at the same time, empower the team to continue the work without dependency on a senior developer.
0:57: Yeah, because remember, it was a tight budget and a tight timeline. Okay, so here was my key move. I did a proof of concept.
1:07: Yes, DCMS, okay, expose a single endpoint that returned the site structure as a folder tree. Okay, it was basically on every page organized inside folders.
1:21: For the tree, basically. So I built quickly a proof of concept to show that it was possible. I grabbed the tree structure, pull, page ID, page size, and confirm.
1:33: We could process the content easily. I just rushed for that. Yeah, easily. I used the recursive programming. There was no charity at that time.
1:43: in the study. I designed a migration pipeline that was simple enough for a junior team to run safely because if I had to suppose a Drupal migration API, Drupal OG or Indeprogramming, all the service stuff and so on, there was too much, you know, for a junior team that it was just starting.
2:03: Okay, so instead of doing everything through the Drupal migration API, which could add complexity, the time I was on a learning curve, as I mentioned, I used Drash commands to process content into batches.
2:16: I created a simple table with a page ID, page type, a success status, and a blank and triple entity ID column.
2:26: Yeah, so the process was repeatable. Fetch page IDs, store them, process them, batches, map the fields, create the nodes. Easy!
2:36: Marks assist and store the Drupal entity ID. Yes, this avoid timeouts and allow save reruns anytime. Yeah. But what made this a real success wasn’t only the code because of the budget and timeline.
2:54: Yeah, like right now, I had to use the clients existing resources, the junior developers and empower them as soon as possible.
3:04: Fast. So I built a simple UML class diagram showing the services that flow the field mapping examples, text fields, flow as entity reference to paragraphs, documentation, tick and assignment, and I can’t style plan so that the PMs can also understand all the stakeholders plus leaf training sessions.
3:25: And of course, Lombidio. This made immigration not only fast but sustainable. They were able to continue the work confidently without me being a bottleneck.
3:38: Yeah, because very specific at all. In the end, we immigrated successfully under the real constraints. Tight budget, tight deadline, junior team, and leaf editors publishing content anytime.
3:52: The client was extremely happy, and they even left our recommendation in my LinkedIn. in my triple skills were outstanding. The plan was easy to follow and that they work with me again.
4:04: Yay. So the lesson here, it is enterprise delivery isn’t about perfect conditions. Not perfect, as you could see, like my daughter here.
4:15: It’s about the right stride under pressure. If you want, I can record part two, showing the exact pipeline, fetch, store, for the rush process and mapping.
4:27: Ta-da!
