This page provides a comprehensive view of monitoring capabilities, structured into two
key sections: monitoring using dashboard in CAS user interface and metric collection
through Prometheus.
Monitor from CAS dashboard
Go to page of the IBM Fusion user interface to view the CAS dashboard. It provides an inventory of
domains and data sources and newly created domain entries. You can use this information to get
insights of system activity.
CAS metrics
CAS provides a set of Prometheus metrics
to give you insights on processing time of a batch of documents with respect to the document
sizes.
| Metric |
Description |
cas_documents_batch_duration_seconds [label: document_processor_name,
event_type] |
Overall time taken by CAS to
successfully process a batch of new, updated, or deleted documents. |
cas_ingest_service_batch_duration_seconds [label:
document_processor_name] |
Time taken by NVIDIA ingestion service
to successfully process a batch of new or updated documents. This is a subset of the overtime time
measured by cas_documents_batch_duration_seconds. |
cas_documents_size_bytes_total [label: document_processor_name,
event_type] |
Total size of all documents in a batch of new, updated, or deleted documents. This metric
is published regardless of whether the batch processing is successfully or not. It gives you
insights on whether there is any unexpected spike of document size. |
cas_document_size_bytes [label: document_processor_name,
event_type] |
Size of each document. This metric is published regardless of whether the batch processing
is successfully or not. It gives you insights on whether there is any unexpected large
file. |
The following table provides label definitions:
| Label |
Value |
document_processor_name |
domain name |
event_type |
update: new or updated documents |
| |
delete: deleted documents |