Troubleshooting Meltdown

Troubleshooting Meltdown

Debug page

If NODE_ENV is not production the developer can view /debug displaying a summary of pages in the site and info about each.

Unmemoized nodes

If you see an error in the console such as:

Trace: detected unmemoized nodes in [
  <ref *1> SiteClient {

There might be an issue where the content of a tag is null or undefined.

Ways to troubleshoot:

Comment out some HTML tags you recently added.
Debug the values of dynamic values e.g.
"val: {JSON.stringify(val)}"
Search for browser-only variables window.* or navigator.*
Enable imba --inspect in the script npm run dev-server in package.json

It can be a difficult challenge to trace down precisely what is causing it. We can work with the imba team to make it easier in the future.

Server keeps restarting

There might be some files that are frequently changed thus causing a server restart. The imba tooling uses the ignore module so something you can try is adding the files that are changing to .gitignore.

Unexpected end of file in JSON

An educated guess is that when Meltdown is writing files to disk, and while they are still being written, imba+esbuild are trying to read and parse them before it's done.

The console error looks like:

✘ [ERROR] Unexpected end of file in JSON

We'll work together with the community to figure out a solution. For the time being it might just print the error and try again and succeed.

Test for broken links

Search indexer

# the output includes 404 and 500 errors
npm run search-index

Linkinator ⚠️ At this time it wont obey --include and it will spider all links, including some or all external links.

npm install linkinator

# default site port
./node_modules/.bin/linkinator http://127.0.0.1:33765

wget

wget --spider -o wget-spider.log -e robots=off -r -p http://127.0.0.1:33765

# missing
cat wget-spider.log | grep ' 404 '

# errors
cat wget-spider.log | grep ' 500 '