Hi, @jirevwe! Would love to get some help on this:

Running this, what I'm seeing is all the containers come up, except one — the web container. The error I am seeing is this:

Seems to be an issue with mongo replication!? I did try increasing the connection timeout to 30 seconds and replicating settings similar to the replica-set.sh file but that did not resolve this. Perhaps mounting the replica-set.sh onto the servers would solve this?

Do let me know how it's looking and what changes I can make to make Uffizzi previews valuable for convoy!

Hey @jirevwe

Friendly ping - wondering if you were able to take a look at this? Would like to fix this and make Uffizzi previews serviceable and efficient for Convoy. Let me know :)

Hi @ShrutiC-git, taking a look at the logs does seem like an issue with the mongo replica set. You can take a look at this file docker-compose.yml to see how we currently recommend setting up a Mongo replica set with Convoy along with an example of what the convoy.json should look like.

thanks, @Dotunj, taking a look ~

hey @Dotunj, thanks again for your help on this. The example you sent was very helpful!

cc @jirevwe
Here is the PoC PR which creates a preview environment. This is the comment on that PR, pointing me to the preview.

So, all the containers do come up, and looking at the logs, it seems the web container is now starting successfully. I see this line in the log - {"level":"info","msg":"Server running on port 5005","source":"api server","time":"2023-02-03 21:18:26"}. Also, it seems to be receiving requests from the browser - which is great.

The worker and the scheduler also are up and running.
I see these in the logs, indicating successful start of these container as well:

• worker → {"level":"info","msg":"Worker running on port 5006","source":"worker","time":"2023-02-03 21:18:26"}
• scheduler → {"level":"info","msg":"Scheduler starting","source":"scheduler","time":"2023-02-03 21:18:24"}

Also, redis, mongo (the replica set) and mongo-setup (replica setup) is running successfully too.

What is weird is I would randomly see a 503 error on the preview.

It'd load the sign-in screen. I'd log into the preview using the username [email protected] and password default. It would log in, but at times, it'd give a 503, indicating that either the web service or upstream service is unavailable. However, the logs do not indicate that anything is wrong.

The same request on retrying would serve the page successfully. Seeing this behavior also on the sign-in screen. Could it be a memory issue? The memory I have allotted to the services is rather on the higher side, but maybe it requires more.

Since I'm not seeing this error constantly, The same request on refreshing serves the page, and vice-versa ( if the sign-in screen is serviceable, on refreshing, it'd throw a 503). This does indicate that it is not an issue with the configuration itself, but perhaps the web service or the upstream service - scheduler or worker is getting killed — but the logs should indicate this. However, the logs look clean.

Not sure what could be the issue. Could taking a look at the PoC PR help? How could we debug this more efficiently?

Super glad we moved a few steps closer - the preview does come up and the containers seem healthy. Nothing obvious I could point to. Looking forward to making the final changes and having the previews serviceable and efficient!

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}