Hi,
I encounter a problem with the API key management in Kibana.
The page behind Stack Management -> Security -> API keys does not load. After some time it displays "Could not load API keys."
Under technical details is a "proxy error".
I tried to directly call the API endpoint /_security/api_key in the developer console but get this response:
When I search in the Kibana logfiles I only find one mention of api_key:
grep -e "api_key" /var/log/kibana/kibana.log
{"service":{"node":{"roles":["background_tasks","ui"]}},"ecs":{"version":"8.6.1"},"@timestamp":"2024-01-18T05:17:30.983+01:00","message":"Failed to mark Task alerts_invalidate_api_keys \"Alerts-alerts_invalidate_api_keys\" as running: Request timed out","log":{"level":"ERROR","logger":"plugins.taskManager"},"process":{"pid":23200},"trace":{"id":"20fb114172fcef1bf344bc14f8f1dd25"},"transaction":{"id":"092958f692f4f72b"}}
It is some time ago I was on the API key page so I can't tell the version, it broke. Currently we run 8.12.0.
I have increased the timeout for calls from Kibana to Elasticsearch to 120 seconds instead of the default (30).
Now calling GET /_security/api_key in the developer console results in this error:
{
"error": {
"root_cause": [
{
"type": "illegal_argument_exception",
"reason": "RecyclerBytesStreamOutput cannot hold more than 2GB of data"
}
],
"type": "illegal_argument_exception",
"reason": "RecyclerBytesStreamOutput cannot hold more than 2GB of data",
"suppressed": [
{
"type": "illegal_state_exception",
"reason": "Failed to close the XContentBuilder",
"caused_by": {
"type": "i_o_exception",
"reason": "Unclosed object or array found"
}
}
]
},
"status": 400
}
There are many such errors in search engines. I'm trying to figure out, what the cause could be. I only want to load the API keys. That should never generate 2 GB of data.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.