dse.metrics - Performance Metrics

class Metrics

A collection of timers and counters for various performance metrics.

Attributes

request_timer

= None

A greplin.scales.PmfStat timer for requests. This is a dict-like object with the following keys:

  • count - number of requests that have been timed

  • min - min latency

  • max - max latency

  • mean - mean latency

  • stdev - standard deviation for latencies

  • median - median latency

  • 75percentile - 75th percentile latencies

  • 97percentile - 97th percentile latencies

  • 98percentile - 98th percentile latencies

  • 99percentile - 99th percentile latencies

  • 999percentile - 99.9th percentile latencies

connection_errors

= None

A greplin.scales.IntStat count of the number of times that a request to a Cassandra node has failed due to a connection problem.

write_timeouts

= None

A greplin.scales.IntStat count of write requests that resulted in a timeout.

read_timeouts

= None

A greplin.scales.IntStat count of read requests that resulted in a timeout.

unavailables

= None

A greplin.scales.IntStat count of write or read requests that failed due to an insufficient number of replicas being alive to meet the requested ConsistencyLevel.

other_errors

= None

A greplin.scales.IntStat count of all other request failures, including failures caused by invalid requests, bootstrapping nodes, overloaded nodes, etc.

retries

= None

A greplin.scales.IntStat count of the number of times a request was retried based on the RetryPolicy decision.

ignores

= None

A greplin.scales.IntStat count of the number of times a failed request was ignored based on the RetryPolicy decision.

known_hosts

= None

A greplin.scales.IntStat count of the number of nodes in the cluster that the driver is aware of, regardless of whether any connections are opened to those nodes.

connected_to

= None

A greplin.scales.IntStat count of the number of nodes that the driver currently has at least one connection open to.

open_connections

= None

A greplin.scales.IntStat count of the number connections the driver currently has open.

Methods

get_stats

()

Returns the metrics for the registered cluster instance.

set_stats_name

(stats_name)

Set the metrics stats name. The stats_name is a string used to access the metris through scales: scales.getStats()[<stats_name>] Default is ‘dse-<num>’.