AxonOps Application Dashboard Metrics Mapping¶
This document maps the metrics used in the AxonOps Application dashboard to their corresponding sources.
Dashboard Overview¶
The Application dashboard provides visibility into client application interactions with Cassandra, including throughput metrics (reads, writes, batches) and connection information. It helps monitor application-level performance and usage patterns.
Metrics Mapping¶
Throughput Metrics¶
| Dashboard Metric | Source | Description | Attributes |
|---|---|---|---|
client_throughput_read |
Client metrics | Read operations throughput | username, remoteIP, keyspace, table, dc, rack, host_id |
client_throughput_write |
Client metrics | Write operations throughput | username, remoteIP, keyspace, table, dc, rack, host_id |
client_throughput_batch_logged |
Client metrics | Logged batch operations | username, remoteIP, keyspace, table, dc, rack, host_id |
client_throughput_batch_counter |
Client metrics | Counter batch operations | username, remoteIP, keyspace, table, dc, rack, host_id |
client_throughput_batch_unlogged |
Client metrics | Unlogged batch operations | username, remoteIP, keyspace, table, dc, rack, host_id |
Connection Metrics¶
| Dashboard Metric | Description | Attributes |
|---|---|---|
cas_Client_connectedNativeClients |
Number of connected native protocol clients | dc, rack, host_id |
cas_authentication_success |
Successful authentication attempts | username, dc, rack, host_id |
Query Examples¶
Reads per Second¶
sum by(remoteIP, username, keyspace, table) (client_throughput_read{axonfunction='rate',dc=~'$dc',rack=~'$rack',host_id=~'$host_id',keyspace=~'$keyspace', table=~'$scope'})
Writes per Second¶
sum by (remoteIP, username, keyspace, table) (client_throughput_write{axonfunction='rate',dc=~'$dc',rack=~'$rack',host_id=~'$host_id',keyspace=~'$keyspace', table=~'$scope'})
Logged Batches per Second¶
client_throughput_batch_logged{axonfunction='rate',dc=~'$dc',rack=~'$rack',host_id=~'$host_id',keyspace=~'$keyspace', table=~'$scope'}
Native Connections¶
ceil(cas_Client_connectedNativeClients{dc=~'$dc',rack=~'$rack',host_id=~'$host_id'})
Successful Authentications Rate¶
sum(cas_authentication_success{axonfunction='rate',dc=~'$dc',rack=~'$rack',host_id=~'$host_id'}) by (username)
Panel Types and Descriptions¶
Throughput Section¶
-
Reads/sec - Line chart showing read operations per second by user, IP, keyspace, and table
-
Writes/sec - Line chart showing write operations per second by user, IP, keyspace, and table
-
Batches/sec (logged) - Line chart showing logged batch operations per second
-
Batches/sec (counter) - Line chart showing counter batch operations per second
-
Batches/sec (unlogged) - Line chart showing unlogged batch operations per second
Connections Section¶
-
Native connections - Line chart showing number of connected native protocol clients
-
Successful Authentications by user (rate) - Line chart showing authentication success rate by username
Filters¶
-
data center (
dc) - Filter by data center -
rack - Filter by rack
-
node (
host_id) - Filter by specific node -
username - Filter by client username
-
keyspace - Filter by keyspace
-
table (
scope) - Filter by table
Metric Details¶
Client Throughput Metrics¶
- Track operations at the application level
- Include client identity (username, IP address)
- Provide keyspace and table level granularity
- Use
axonfunction='rate'to calculate per-second rates
Connection Metrics¶
connectedNativeClientsshows current connection countauthentication_successtracks login attempts- Both metrics help monitor client access patterns
Legend Format¶
The dashboard uses descriptive legends combining multiple attributes:
- Throughput:
$username @ $remoteIP $keyspace $table - Connections:
$dc - $host_id - Authentication:
$username
Notes¶
- Client throughput metrics provide application-level visibility not available in standard Cassandra metrics
- These metrics help identify heavy users, problematic applications, or unusual access patterns
- The
ceil()function is used for connection counts to ensure whole numbers - Batch type separation (logged/unlogged/counter) helps monitor different write patterns
- Remote IP tracking enables geographic or network-based analysis of client access