Sprint 3 is now completed and it is time for the final retrospective. Although this sprint was shorter than the others, my team and I made good progress on our substem. With two other sprints under our belts, my team was able to complete what we needed to so everything would be wrapped up and ready for next year’s class to work on it.

A change we made that helped the team was breaking down our issues better as well as working together to tackle those issues. Our goal for this sprint was to have a completely working test so we broke down the process of creating the test into several issues. With everyone working together on the code, we were able to get through each issue quicker than we would have as individuals. We still faced challenges but being able to rely on the other members of the group was a big help.

Something that my team could improve on is understanding docker and using the servers. The most difficult problem we encountered was counting with the server. Our first few issues were not too difficult and we were able to solve them relatively quickly but getting the test to run automatically by connecting to the back-end server was difficult. We first ran into issues just connecting, then we had issues with the server timing out. We relied heavily on the guest info system group for help with this. Docker and the severs are a huge piece of this project so a better understanding of how to get everything working would have been a huge help to the team.

In this sprint, we only worked on the backend since our object was to get an automated test up and running. Our issues were mostly worked on as a group. The following are the issues that I was involved with for this sprint:

One of the easier issues we had was reviewing the current API:

https://gitlab.com/LibreFoodPantry/client-solutions/theas-pantry/inventorysystem/inventoryapi/-/issues/8

Creating the getInventory manual test was another one of our issues:

https://gitlab.com/LibreFoodPantry/client-solutions/theas-pantry/inventorysystem/backend/-/issues/54

Another issue we worked on was fixing the get inventory returning a string bug:

https://gitlab.com/LibreFoodPantry/client-solutions/theas-pantry/inventorysystem/backend/-/issues/60

As an individual, I believe I could improve by working on my collaboration. Even though my team asked for help, I personally did not. During this sprint, I relied on my teams to go and ask questions to the other groups. I think as an individual I could take more responsibility when it comes to collaborating with other teams.

I think sprint 3 was another positive experience for the team. We continued to adapt and make the changes we needed to that we talked about from our previous sprints. This allowed us to better ourselves as developers and get through the issues we decided to work on. Even though this is the end of time together working on LibreFoodPantry, we will take the lessons we learned with us as we start our careers. 

From the blog CS@Worcester – Ryan Klenk's Blog by Ryan Klenk and used with permission of the author. All other rights reserved by the author.

The third sprint for my software development capstone class was on an honest note a bumpy road. The purpose of the sprint was to start the next level implementation of creating test files using mocha and chai and also making sure the API is validated and can be bundled properly. Through the first part of the sprint, the team has been very productive in keeping up with the issues we made for ourselves. After Spring break, our productivity started to plummet since we did not accomplish issues the following, we did not have our class meetings.

As for my team members and I, we have faced great issues that we started to come across that stumbled all of us. Due to facts of, lack of knowledge, merge requests, and was not able to accomplish all our issues for the sprint. First, when it came to merging requests, I admit we have gotten better at how Git operates. But when it came to approving the request assigned to us. We had a pile of request that was not merged and had to spend our class meetings squashing all of the time to keep the main repository up to date as soon as possible. Secondly, our productivity has plummeted since our spring break, we did not have to work on anything which was completely optional. However, the following week there was no class reunion and I made no effort and treated it like another extension of spring break. Third, as I gained a deeper understanding of the problem, I didn’t know how to deal with some of the issues. The other issues couldn’t be resolved and were interrelated, so we had to address them in the next final sprint.

Our team decided not to undermine our productivity. Especially if the third sprint is the last race in the podium category. We decided and wanted to be more active and help each other in the workspaces when there are things we don’t know or understand. Talk about gathering information on all the advice our professors can give us to put us on the same page.

As for me, I have done good things, but of course, I could have made better improvements and made a better contribution to our team. I wanted to know more about how the Mocha and Chai tests were implemented. However, I have decided to do more key pieces to solve the problem. However, in the next problem, we will use one or more of the problems we created to create a more robust test file that will pass all tests. Talk to team members who have done an extensive research to see if they can help you with modifying and implementing test files that have already been created. At least the experience provided.

Sprint 3 Issues

• Added bin commands to the ReadMe on the backend repository. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/issues/33
• Composed a meeting with the IAM team for their token format. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeperapi/-/issues/11
• Composed a meeting with The AWS team. notified that the team is are not ready to deploy. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/issues/23
• Made modifications to the ReadMe in the documentation repository. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/documentation/-/issues/1
• made sure the test pass for API returning endpoint using mocha and chai testing. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/issues/25
• Made an endpoint to the backend repository for returning the API version. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/issues/13
• Made a data file for the endpoint of API return. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/issues/12
• made test file for returning API endpoint using mocha and chai. https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/issues/24

From the blog cs@worcester – Dahwal Dev by Dahwal Charles and used with permission of the author. All other rights reserved by the author.

Here in sprint 3, things seemed to have progressed smoothly with the project in that we have finished most of what we have added in the product log although there were some areas of difficulty. The cohesion of the group has improved with better communication on what is to be done and the planning together seems to have gotten better. There are still improvements to be made but overall, the project is in a good state right now. The presentation also seems to be going well and should be alright. The quality of the sprint overall seems to be up and is providing good results.

On what has worked well, we have managed to get to the end of the semester and be able to finish a good amount of work on the project. We added more calls to the backend and continuously updated the API with new changes. The backend is set up with the proper semantic versioning. The android nest tester is functional and everything that we have built looks to be solid. The communication in the group looks to have improved as people communicated on what they did and were working on more than in previous sprints. The team also seemed to have gotten together more often to work on issues. The group stuck to what was planned pretty well and followed the timeline agreed upon. The issue board moved along more than the last sprint, issues were moved across the board compared to last sprint where nothing was moved until the last day.

On what has not worked well, there was unfinished work at the end of the sprint that now has to be sent over to next semester. This means that we have made mistakes during our planning on the difficulty of the work on the issues. The planning of getting the presentation done also could have been better. There probably should have been more dedicated time to get together and figure out the presentation as it felt a bit odd at the end figuring out what topics to do.

We could improve as a team by firstly, we could have worked together more regarding the issues regarding key authentication that we were unable to implement. There could have been chances to get together and try to figure out what was wrong and get it done by the end of the sprint. We could also improve as a team in our planning and plan more effectively on what can be done this sprint. We also could have been more planning with the presentation and knowing what exactly what portion each member was contributing with.

I could improve personally by putting in more of my input and getting my ideas out there. I also could have contributed more on different issues within the project. I also could have done better with working together with the group members such as when we were working on implementing the range call where it felt like backseat coding and I could have contributed more.

From the blog CS@Worcester – kbcoding by kennybui986 and used with permission of the author. All other rights reserved by the author.

This final sprint was much more of a struggle than the other two prior sprints. Since we got a lot of progress done in achieving our goals throughout the past two sprints, we figured that this sprint would let us bring everything together and finalize everything we have worked on up until this point. Unfortunately due to unforeseen issues, we were unable to reach all of our goals, an we were instead stuck on one task for a vast majority of the sprint. Our goal this sprint was to test the backend and the endpoints to ensure that they can all work with each other in the way that we planned. We initially thought that we had solved many of the issues with the backend that have been plaguing us in the past sprints, but unfortunately that was not the case. In order to test the endpoints we needed to be able to run the backend, and when we attempted to do that we realized that what the previous group had left us was not configured correctly and was also missing a lot of important code. Upon realizing this, Kurt and I went to work trying to figure out what exactly was missing and preventing the backend from running correctly. This debugging took up a vast majority of the time, and in the end Kurt was able to figure out what the main issues were and managed to get the backend partially running. This was still not enough for us to get our endpoints tested however.

Given that this sprint was more of a challenge than initially expected, there are definitely things that we could have done better as a group in hindsight. One of the bigger mistakes I believe we made was that we didnt have everyone working on the backend problem. Since this was the priority, it should have been a team effort to try to solve it and get it running as fast as possible so that we can test the rest of our code. Unfortunately since in the beginning we didnt realize the scope of the issue we never organized the work in that way. Many of the team members were forced to wait until the backend was complete before they could try to do the tasks that they were assigned. However since the backend never ended up getting fully completed these team members did not have as many opportunities to contribute.

Given hindsight about my performance in this sprint, and knowing what the issues we faced were I can look back and see that there were definitely things I could have done better. One of the main things that comes to mind is my initial approach to trying to solve the backend issue. At first I simply attempted to brute force the problem, moving code and files I thought might be affecting the issue around in the project to see if that changed affected anything. This proved to be a useless exercise in the end and resulted in me wasting more time than I should have. If I could do it again I would approach the issue more slowly, reading through all the relevant code and researching what issues may cause the kinds of problems that we were experiencing.

I do not have any gitlab links to share, since I was unable to get the backend working and therefore had nothing to push. As I mentioned a majority of my work was brute force trial an error where I attempted to figure out the issue.

From the blog CS@Worcester – Sebastian's CS Blog by sserafin1 and used with permission of the author. All other rights reserved by the author.

For this sprint, my team, team 1, managed to complete all the epics that we had, which were .gitlab-ci.yml, fixing API, modifying backend, finishing frontend, and completing documentation project. I believe my team is done with the API, backend, and frontend. A small portion of the gitlab-ci.yml and the documentation project are still in progress.

In this sprint, I have worked with backend and frontend projects. For the backend, first of all, I modified the API to reduce the number of endpoints from 5 to 2 which getReport and getVesion. Then, to send fake data without using any endpoints, I created a folder named send_Fakedata_RMQ. This folder contains two files, one is used to contain dummy data and the other one is used to create rabbitmq connection and send those dummy data to a rabbitmq channel. So, whenever we use the command of node ./src/send_Fakedata_RMQ/sendData.js, we are sending dummy data to rabbitmq. These data are used to test how the entire backend works. To automatically retrieve those data from rabbitmq and store them into mongodb, in the index file, I added a new method called RabbitMQ_service(). In this method, I used setInterval() function to repeatedly retrieving data from rabbitmq every day. The fixed time used to delay between each call should be converted from day to ms, 86400000.

For the getReport endpoint, I’ve added two new properties called new_household and new_people, which represent the two columns of the Newsince field in the example report. Furthermore, I also edited the csv.js file to make the getReport endpoint return the report in text form instead of downloading it directly to the computer.

For testing, I created two tests for this backend. One is used to test the response status, the other one is used to test the response data. To test the response data, I added a new file called expectedResult.js in the testing/test/lib folder. This file contains a string where each value in this string is manually counted based on dummy data I sent to Rabbitmq. This string is supposed to be my expected result which I am using to compare with the actual result recieved from the system.

One of the problems of the backend is about what should be stored in the guestInfo collection. Since I misunderstood the properties of the guest variable in json, I stored each guest as a unique guest in the collection without the “date” property. The requirement of this collection is that every record of each guest must be stored with a date and time, so this issue must be fixed in order to produce an accurate report in the near future.

For frontend, I added the same folder named send_Fakedata_RMQ from the backend to send dummy data to check whether the frontend is working or not. In getReport.vue, I added a method called downloadCSVfile(data) to download a file named WCFB_report.csv, when the “Submit” button is clicked. I also added a simple instruction of how to run the frontend in README.md. I believed we are done with the frontend and it works fine as expected.

For this sprint, we all focused and worked on different topics to perfect the reporting system as much as possible. Most of the time we worked well as a team, but there were still some points that I think we should improve in the future. Like what I just said, each team member worked on a different topic, so it is hard for us to keep track of all topics at the same time. Team members should actively share the problems they are facing so that the whole team can look at it and help them as soon as possible. Don’t stay silent and wait until the last minute without giving any reason to the team. I mean each team member should be aware of his/her responsibilities to the whole team. On the other hand, for myself, I think I should learn more about teamwork skills to ensure that every team plan is completed by every member before the deadline; or at least I can come up with some solutions to deal with similar situations.

From the blog CS@Worcester – T's CSblog by tyahhhh and used with permission of the author. All other rights reserved by the author.

In our final sprint, our team focused primarily on finishing our demo, communicating with other teams about their projects and how they will fit into the cluster, and creating a schematic of the final cluster. I think our final sprint went really well. While we did not finish everything we set out to, we were able to get a significant amount of work done, and I think we have left the project in a good place for whoever will be picking it up next.

I think a lot of things went well in this sprint. Similar to our second sprint, we had a much clearer understanding of what we wanted to accomplish in this sprint, and that made it much easier to divide work into manageable tasks. Unlike the last sprint, much of this work could happen concurrently, so team members were blocked less often by the issues other team members were working on. Additionally, I think we were all much more comfortable communicating with each other, and this allowed our work to go much more smoothly.

The only thing that didn’t go well was how much work we wound up getting done by the end of the sprint. We initially planned for thirty points of work, and by the time the sprint had officially ended we only got through about half of that. I think we underestimated how long some of our issues would take to complete and wound up falling behind. Regardless, I feel that we got a substantial amount of work done.

Our team worked very well during this sprint, but there is always room for improvement. I think we could improve on our time management a little bit. We spent a lot of time discussing how we would like the sprint to go, and this left less time to actually work on our issues. Still, I don’t think this impacted us too heavily; I am happy with the work we did get done.

Personally, I think I could improve on the way I write documentation, especially summaries of research I’ve done. I tend to get overwhelmed by whatever information I’m reading and will forgo some details in my retelling. I want to be able to write more detailed documentation since that would be more beneficial for my team members.

Contributions:

Research Helm – I spent time looking into Helm, which is a package manager for Kubernetes. I researched what it is, how it works, and whether it would be worth using in the final cluster for Thea’s Food Pantry.

Communicate with Team 1: Inventory System – I informed the Inventory System team about our work, and I made sure we understood how their project is structured.

Communicate with Team 3: Food Keeper – This is similar to the previous issue. I informed the Food Keeper team about our work, and I made sure we understood how their project is structured.

Communicate with Team 4: nest Guest Info – This is similar to the previous two issues. I informed the Nest Guest Info System team about our work, and I made sure we understood how their project is structured.

From the blog CS@Worcester – Ciampa's Computer Science Blog by robiciampa and used with permission of the author. All other rights reserved by the author.

Our third sprint was definitely the toughest, and we ran into some issues throughout the sprint that really slowed us down. We started out at fixing problems for the project from the end of the previous sprint, then moved on to the last issues we put in the backlog. One issue in particular would take a lot of time to solve and prevented some of the other issues from being completed.

Starting with what worked well, we were doing better at communicating with each other and finished with enough weight for the review. When we were having problems with the issues, we were keeping in touch with each other, and most of us were working together on the main issue that was holding us back. Some of us were having more success than others, but it was eventually completed and the other issues that were relying on it were being finished too. We were also able to finish the presentation on time, and had a plan for putting the presentation together with slides and an outline that we put together.

Our main problem for this sprint wasn’t a problem with the team, but with one of the issues we were working on, which had to do with building the backend server and getting it working for everyone. Along with getting the server working, some of us had different issues that others weren’t getting. We spent a lot of time working on this issue, and had to rush through the last issues in the last week. Aside from that, we were still trying to improving on keeping up the merge requests up to date. We managed to get all the merge requests done in time, the requests would just build up over a few days. Aside from that we had some days where team members would not be there in person, but they would join online and were able to join in the in-person discussions through the discord chat.

I think I did well at continuing with the improvements I was making as a team member by communicating with the team when working on the project and the presentation, but there were a couple times when working on the presentation that I missed out on some discussion. I would later join in the discussions, catch up on what I missed, and do my part in recording for the presentation. When waiting on the backend server issue being completed, I had the idea of setting the api tests up with a template for the test files, so that they would be mostly done by the time we are able to use the api. That way, the issues could still be completed while dealing with the backend server. I got this idea from a template file that we had made in the previous sprint for the data files, and a template file for api testing that was created during this sprint, and figured we could have something similar for the api testing methods while waiting. In the endpoint api templates, the testing methods were set up for the api to accept sample valid and invalid calls that worked for samples we had, so when we would be able to use the api for the testing, we could just make new calls to put in those positions and then adjust based on testing results.

https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/merge_requests/28

Resolving the problems with the category test file so all tests passed.

https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/merge_requests/33

Template for the category api testing that was created while waiting for the backend server.

https://gitlab.com/LibreFoodPantry/common-services/foodkeeper/foodkeeper-newbackend/-/merge_requests/36

Template for the product api testing that was created while waiting for the backend server.

From the blog Jeffery Neal's Blog by jneal44 and used with permission of the author. All other rights reserved by the author.