We are preparing to deploy our new app with cleavr, and I have few questions that popped up when I was testing the platform, and maybe can offer some suggestions for the future.
- Builds and github actions:
- Are there plans to introduce live build logs? Right know they are not available during the build time, and if something hangs up, no one knows about it, and this leads to timeout. We are building from github now, but this could be an issue especially for smaller servers which are strained by the installation.
- If github build is enabled, applicaton env variables are not propagated to the build process - in our case nuxt BASE_URL was not set after build, maybe we have misconfigured that? Editing github workflow file and filling it with env solved that issue, and I still think that this feature works as intended, am I right? Some documentation would be appreciated nonetheless
- While testing staging env, we spun up t2.small aws instance, and after few deployments, it got clogged with deployment data, in effect bricking up the machine. I saw cleaning old deployment action, but I don’t see that it deletes old deployment files, what does it do in essence? And what can we do to limit deployment history to some arbitrary number? For example staging env doesn’t even need deployment history, we will just redeploy a new version if something will fail.
- Running the application with reporting
- We have some monitoring infrastructure set up for our services, namely datadog, and we would like to report app request logs to the service. Right now, even with pm2 I couldn’t grok whether there are some kind of server logs from the nuxt server by which I mean server output to stdout or file, because the standard ones, beside having information about the app start, were empty. Could you help me with that?
Other than that, the setup was a breeze and we enjoyed the platform a lot, thank you for help!