Massive performance degradation on indexing after upgrading to 1.13 from 1.8

TL;DR

What was reported: Issue meilisearch/meilisearch#5616 (closed, 29 comments): Describe the bug I've upgraded from 1.8 to 1.14, and hit a really nasty performance issue, which I can only contribute to the newer versions, not my own indexes, as nothing changed and I never seen this before.

Steps

1. What was reported:

Issue meilisearch/meilisearch#5616 (closed, 29 comments): Describe the bug I've upgraded from 1.8 to 1.14, and hit a really nasty performance issue, which I can only contribute to the newer versions, not my own indexes, as nothing changed and I never seen this before. My memory consumption is very low, under 2GB normally and was the case on re-deploys as well. But now, when I deploy the instance, it hits 4GB of memory and then goes to under 1GB. This is a huge problem, as I am unable to deploy to a smaller instance, because it just fails when going over the memory (let's say 2GB).

2. What works:

Massive indexing memory spike after upgrading Meilisearch (1.8 to 1.13/1.14) inside Docker: the new indexer does not detect container memory restrictions and uses host memory limits. Fix: cap indexing memory with the env var MEILIMAXINDEXINGMEMORY (e.g. MEILIMAXINDEXINGMEMORY=30GiB), sized to your container. Also upgrade to the latest Meilisearch release, which includes further indexing memory improvements.

When to use

You are seeing this: My memory consumption is very low, under 2GB normally and was the case on re-deploys as well. Use this skill when you run into "Massive performance degradation on indexing after upgrading to 1.13 from 1.8".

When not to use

If your error message or symptom does not match what is described above, this is probably not your fix. Search for your exact error text instead of forcing this one to fit.

Versions

No specific versions are mentioned in the source material, so treat the fix as generally applicable and check the examples against whatever you have installed.

Why this happens

The original report does not dig into a root cause. It documents the symptom and the fix that resolved it.

Edge cases

• Also upgrade to the latest Meilisearch release, which includes further indexing memory improvements.

In the GitHub issue, the reporter asked: Massive performance degradation on indexing after upgrading to 1.13 from 1.8. The working answer that came out of the thread: Massive indexing memory spike after upgrading Meilisearch (1.8 to 1.13/1.14) inside Docker: the new indexer does not detect container memory restrictions and uses host memory limits. Fix: cap indexing memory with the env var MEILIMAXINDEXINGMEMORY (e.g. MEILIMAXINDEXINGMEMORY=30GiB), sized to your container. Also upgrade to the latest Meilisearch release, which includes further indexing memory improvements. Source: https://github.com/meilisearch/meilisearch/issues/5616