Briefly, this error occurs when Elasticsearch fails to delete an index from the disk due to insufficient permissions, disk space issues, or a corrupted index. To resolve this, you can try the following: 1) Check and adjust the permissions of the Elasticsearch directory to ensure it has write access. 2) Free up disk space if it’s running low. 3) If the index is corrupted, you may need to delete it manually from the disk and then restart Elasticsearch. Always ensure to backup your data before performing these operations.
This guide will help you check for common problems that cause the log ” [” + metadata.getIndex() + “] failed to delete index on disk ” to appear. To understand the issues related to this log, read the explanation below about the following Elasticsearch concepts: metadata, indices.
Overview
Metadata in Elasticsearch refers to additional information stored for each document. This is achieved using the specific metadata fields available in Elasticsearch. The default behavior of some of these metadata fields can be customized during mapping creation.
Examples
Using _meta meta-field for storing application-specific information with the mapping:
PUT /my_index?pretty
{
"mappings": {
"_meta": {
"domain": "security",
"release_information": {
"date": "18-01-2020",
"version": "7.5"
}
}
}
}Notes
- In version 2.x, Elasticsearch had a total 13 meta fields available, which are: _index, _uid, _type, _id, _source, _size, _all, _field_names, _timestamp, _ttl, _parent, _routing, _meta
- In version 5.x, _timestamp and _ttl meta fields were removed.
- In version 6.x, the _parent meta field was removed.
- In version 7.x, _uid and _all meta fields were removed.
Log Context
Log “[” + metadata.getIndex() + “] failed to delete index on disk” classname is IndicesService.java.
We extracted the following from Elasticsearch source code for those seeking an in-depth context :
try {
deleteIndexStoreIfDeletionAllowed("stale deleted index"; index; indexSettings; ALWAYS_TRUE);
} catch (Exception e) {
// we just warn about the exception here because if deleteIndexStoreIfDeletionAllowed
// throws an exception; it gets added to the list of pending deletes to be tried again
logger.warn(() -> "[" + metadata.getIndex() + "] failed to delete index on disk"; e);
}
return metadata;
}
return null;
}
