Documentation > Wiki > Solr Best Practices

🧠 Solr RAM & Memory Management: Best Practices (or, “How Not to Blow Up Your Server”)

Solr is a beast—it loves RAM like a dog loves a steak. If your Solr server is gobbling up memory and crashing, don’t panic! Here’s what you need to know, plus battle-tested ways to keep things lean, mean, and not out-of-memory.

Why Does Solr Use So Much RAM?

Solr eats memory to build search results, cache data, and keep things fast.
But:

Bad configuration or huge, inefficient requests can cause even the biggest server to choke and burn through RAM.
Sometimes, small indexes on giant machines will still crash if your setup isn’t right.
Good news: Opensolr has self-healing—if Solr crashes, it’ll be back in under a minute. Still, prevention is better than panic.

🔧 Essential Best Practices

1. Save Transfer Bandwidth (and Memory)

Want to save bandwidth and RAM? Read these tips.
Optimizing your queries is a win-win: less data in and out, and less stress on your server.

2. Don’t Ask Solr to Return 10 Million Results

Requesting thousands of docs in one go?
That makes Solr allocate all that data, and cache it, too.
Solution: Keep the rows parameter below 100 for most queries.
Example:
```
&rows=100
```

3. Paginate Responsibly (Or: Don’t Scroll to Infinity)

If you’re paginating over millions of docs (like &start=500000&rows=100), Solr has to allocate a ton of memory for all those results.
Solution: Try to keep start under 50,000 if possible.
The more stored fields you have in your schema, the more RAM will be used for large paginations.

4. Heavy Faceting, Sorting, Highlighting, or Grouping? Use `docValues=true`

Operations like faceting, sorting, highlighting, and grouping can be memory hogs.
Solution: Define your fields with docValues="true" in schema.xml.

Example:

<field name="name" docValues="true" type="text_general" indexed="true" stored="true" />

For highlighting, you may want even more settings:

<field name="description" type="text_general" indexed="true" stored="true" docValues="true" termVectors="true" termPositions="true" termOffsets="true" storeOffsetsWithPositions="true" />

5. Don’t Go Cache-Crazy

Solr caches are great... until they eat all your memory and leave nothing for real work.

The big four:
- filterCache: stores document ID lists for filter queries (fq)
- queryResultCache: stores doc IDs for search results
- documentCache: caches stored field values
- fieldCache: stores all values for a field in memory (dangerous for big fields!)
Solution: Tune these in solrconfig.xml and keep sizes low.