Potential Solutions For Missed API Calls in My Middleware…
0 reactions
Potential Solutions For Missed API Calls in My Middleware Polling Project?
This is my final senior project, and my university pairs students with actual companies to work on solutions to their problems (software).
The following is a brief summary of what we created
This company sells supplies, and we created a middleware that essentially polls their database (new contacts, leads, accounts, sales orders, etc), and we trigger automated emails to email based on whatever criteria.
The way we trigger automated emails is to "tag" a contact programmatically. "Tagging" a contact essentially puts them into an automated workflow, which emails the contact after a certain time (all done on the email service platform)
We have that done and it works. My current issue is how to handle potential request errors (500-level status codes) when calling the external API (to tag a contact). As of now my poller simply moves on to the next contact to tag. My poller logs everything, so you'll see the error.
This is supposed to be automated, so I don't want the company to have to check for any errors manually
My current solution is to log any API request errors separately and have my poller retry them next time it runs (or retry again at the end of its run). My poller runs once a day, and there really isn't a lot of traffic on their database. Also, I do have retry logic covered already (hits the api endpoint after a 500-level status code, a maximum of 3 retries)
#dev #programming #technology
source
This is my final senior project, and my university pairs students with actual companies to work on solutions to their problems (software).
The following is a brief summary of what we created
This company sells supplies, and we created a middleware that essentially polls their database (new contacts, leads, accounts, sales orders, etc), and we trigger automated emails to email based on whatever criteria.
The way we trigger automated emails is to "tag" a contact programmatically. "Tagging" a contact essentially puts them into an automated workflow, which emails the contact after a certain time (all done on the email service platform)
We have that done and it works. My current issue is how to handle potential request errors (500-level status codes) when calling the external API (to tag a contact). As of now my poller simply moves on to the next contact to tag. My poller logs everything, so you'll see the error.
This is supposed to be automated, so I don't want the company to have to check for any errors manually
My current solution is to log any API request errors separately and have my poller retry them next time it runs (or retry again at the end of its run). My poller runs once a day, and there really isn't a lot of traffic on their database. Also, I do have retry logic covered already (hits the api endpoint after a 500-level status code, a maximum of 3 retries)
#dev #programming #technology
source
| 2,000 mlx | |
| 3 | |
| 0 |