Running Docker Jobs with Limited Memory is a practical project for Docker learners processing a small oral-history collection. The provided tool really parses HTML documents and counts Unicode words, but its initial batch exceeds a 160 MiB container memory budget. Finish the whole index by choosing feasible concurrency under the same limit.
One Guided Lab teaches evidence-based diagnosis and complete delivery. One independent Challenge uses a different collection and output layout. A separate two-document sample shows the search experience before you process the assigned sources.
What You Will Learn
- read a container's effective RAM and total RAM/swap limits;
- correlate a failure with the exact container's memory-cgroup OOM record;
- distinguish an exit code or Docker flag from sufficient diagnosis;
- reduce actual HTML batch concurrency without raising the resource budget;
- preserve every source identity, duplicate title and Unicode word count;
- deliver an editable index and explore its real document relationships and downloads.
Who This Course Is For
This course suits beginning Docker users who want to finish a file-processing job within a fixed resource allowance.
Prerequisites: Basic Docker run, terminal paths and bind-mount purpose. The memory options, account IDs, quoted paths and evidence-reading pipeline are explained. Python programming and search-algorithm implementation are unnecessary.
Learning environment: Docker, the indexing image, original HTML documents and read-only Search Desk are prepared. Each lab has an independent collection and container environment.
Frequently Asked Questions
Is the memory failure caused by a synthetic allocation loop?
No. The tool retains real parsed HTML document trees in its batch. A smaller batch lowers that working set while still processing every input. The documents are original synthetic teaching material.
Can I just raise the limit or remove a difficult document?
The task keeps both memory settings at 160 MiB and preserves all assigned HTML. Success means the complete index fits the declared budget, including accented words and distinct documents with the same title.
Does exit code 137 prove an OOM?
No. You retain the actual failed container and match its ID and time to a native memory-cgroup record. The Docker OOMKilled flag can vary on this teaching environment, so the course does not use either value alone as proof.
