Cannot Unload Non-Existent Core — Index Not Ready After Restart

Errors
Solr Error Guide

Cannot Unload Non-Existent Core

An action was attempted on your Opensolr index before it finished loading. This is a temporary, self-resolving error — here's what it means and what to do.


What Happened?

The Error: Cannot unload non-existent core [your_index_name]

Something tried to perform an action (like a reload, commit, or query) on your Opensolr index, but the index wasn't fully loaded into memory yet. Solr looked for it, couldn't find it in its registry, and reported it as "non-existent" — even though the index does exist on disk.


Why Does This Happen?

Think of it like trying to open a program on your computer while it's still booting up. The program is installed — it just hasn't had time to start yet.

Timeline of a Solr Server Restart1Server startsSolr process boots2Indexes loadingOne by one into memoryAction arrives!Index not ready yet = ERROR3Still loading...Large indexes take time4Fully loadedEverything works again

The most common scenarios that trigger this:

Server Restart / Maintenance

The Solr server was restarted for maintenance or an update. While it's starting up, it loads all indexes one by one. If your application sends a request before your index is loaded, you get this error.

Automated Reload / Commit

A scheduled task, cron job, or your application's indexing pipeline tried to commit, reload, or query the index during a restart window. The timing just happened to overlap with the server coming back up.

Replication Sync

On Opensolr's replicated clusters, a replica server might temporarily unload and reload an index during a replication cycle. If a query hits that replica at that exact moment, this error appears.


Is It Serious?

Usually not. This is a transient, self-resolving error.

Once your index finishes loading (typically seconds to a few minutes, depending on index size), everything goes back to normal automatically. No data is lost. No manual action is needed.

No data loss

Self-resolving

No action needed


What Should You Do?

Wait a Moment

If you see this error once or a few times, simply wait. The index is loading. Try your request again in 30-60 seconds. If it works, the error was just a timing issue during startup.

Check Your Error Audit

Open your Opensolr dashboard and go to the Error Audit tab. If you see only one or a few of these errors clustered around the same time, it confirms a restart-related timing issue.

Add Retry Logic

If your application needs to handle this gracefully, add retry logic with a short delay. If a query or commit fails with a 500 error, wait 5-10 seconds and retry. This covers restarts and replication cycles transparently.


When Should You Worry?

This error is only a concern if:

Situation What It Means Action
Error repeats continuously The index may have failed to load entirely Contact Opensolr support
Error persists for 5+ minutes Something may be blocking the index from loading Check your Error Audit for other errors
Happens without any restart Could indicate a corrupted index or config issue Contact Opensolr support

Quick Reference

Item Details
Error Class SolrException
When It Happens During server restart, replication cycle, or index reload — when an action arrives before the index is ready
Severity Low — Transient
Data Loss Risk None
Resolution Wait for index to finish loading (usually seconds). Add retry logic in your application for robustness.

Error Keeps Happening?

If this error persists or repeats every few minutes, something else may be wrong. Check your Error Audit or contact support.