Global indicator

Alerts

On the homage is displayed :

• number of alerts raised on the period in danger/level
• details is shown by management module

DataServer Consumers

Graph. The per type of client (Navigators, Engines, Scheduled Tasks, Risk components) over time sum of response time of EJB calls made to the DataServer.

Components impact on Dataserver

Graph. The sum of response time of remote (EJB/RMI) calls made to the DataServer - per type of components.

Components can be :

• Navigators
• Engines
• Scheduled tasks
• Risk components
• Generic
• Unknown (client that are not monitored)

Usage of grid

Graph. Percentage of calculators that have been used on the period. Same for the calculators still in used at the end of the period.

Engines status

Engines status (running / not running)

Stalled requests

Description of the requests than run since more than 30s.

PS Events

The number ps events sent by dataserver or consumed by clients

Pending events

The size of the different engines queues also called unconsumed events

Event server queue size

The sum of all queues for the event server. Usually close to zero.

Calypso functional and technical indexes

Messages, CRE, Transfers, Trades

over time number of BOMessages, BOCres, BOTransfers, Trades saved or updated in database.

Engines Pending Events

Graph. The per Engine over time number of pending events to be consumed by that engine.

Engines Statistics

For running Engines :

• Number of requests (Engine.process method count)
• The average events processing time
• The weight of the Engine sum of events processing time in relation to all Engines sum of events processing time
• The maximum number of executing parallel threads
• The configured number of maximum parallel threads that can execute or -1
• The number of badevents over the period

These numbers are only valid for the selected time period

Additional metrics (Webview)

For every Engine, a few more metrics

• Number of DBEventLoad
• Size of events pool
• Suspended status of the engine
• Consumed count (not really interesting, reset at each restart)

Clients

Number of connected clients over time.

Calypso API executions

WorkflowRules

Trade Workflow Rules, Message Workflow Rules and Transfer Workflow Rules executed during the selected time period, with :

• The Workflow Rule average response time.
• The Workflow Rule number of executions.
• The weight of the Workflow Rule executions in relation to all selected Workflow Rules executions in term of execution time.

API calls on DataServer side : EJB - RMI over HTTP

Calls made to the DataServer by all clients executed during the selected time period, with :

• The weight of the requests in relation to all others requests executions (average * #calls .
• The average response time of the request.
• The number of executions of the request.

Backend Calls on clients side

At the level of a client (engine server, scheduled tasks…) , EJB/RMI API calls made to the dataserver by that client during the selected time period, with :

• The entry point when available or #NotTagged
• The weight of the requests in relation to all others requests executions (average * #calls .
• The average response time of the request.
• The number of executions of the request.

Trade report

• execution time of the getTradeReport method
• number of trade returned for the getTradeReport

SD Filters

• Number of calls to the SDfilter
• Ratio of accepts regarding the total number of calls to the SDFilter.

SQL Queries and DataBase

SQL Requests

SQL queries executed during the selected time period, with :

• The weight of SQL query executions in relation to all SQL queries executions in term of execution time.
• SQL query average response time.
• SQL query number of executions.
• SQL query average number of rows returned by the request.
• SQL query stalled events.
• SQL query resultset time

Oracle Database details

Available only with additional HPA Infrastructure agents activated and Oracle user permission for statistics. See installation

• tablespace free space
• blocking sessions
• active sessions
• top queries by response time
• efficiency (Oracle statistics indicator)
• resources (Oracle statistics indicator)
• wait class breakdown (Oracle statistics indicator)

Pools usage

All JVM

Per Pool object type during the selected period:

• the minimum available slots in the pool
• the maximum size of the pool
• the maximum size of the queue: maximum of requests waiting to get a free slot to use

A non exhaustive list of monitored pool :

• DS-CalypsoDS: DataServer Pool of DSConnections:
• EJB3-default
• EJB3-remoting
• JCA-long: Java Connector Architecture long running threads.
• JCA-short: Java Connector Architecture short running threads.
• Mdb-mdb-strict-max-pool|CacheUpdateSubscriberBean: Bean pool for CacheUpdateSubscriberBean message driven bean.
• Mdb-mdb-strict-max-pool|CalypsoAuditMessageBean: Bean pool for CalypsoAuditMessageBean message driven bean.
• Mdb-mdb-strict-max-pool|ReadOnlyCacheSynchronizationBean: Bean pool for ReadOnlyCacheSynchronizationBean message driven bean.
• Messaging-ScheduledThreadPoolExecutor: Pool for threads managed by ScheduledThreadPoolExecutor
• Messaging-ThreadPoolExecutor: Pool for threads managed by ThreadPoolExecutor.
• HikariPool: Database connection pool
• TomcatEndPoint-http-nio: HTTP calls pool for Tomcat (v17)
• CalypsoThreadPool: Calypso pool object

Risk Analytic & Jobs

Dispatcher status

Dispatcher status by dispatcher configuration :

• queues size: average and maximum
• calculators usage: total calculators, max busy, currently busy

Jobs activity on dispatchers :

• identification: name, type, priority
• number of waiting jobs: average and maximum
• number of running jobs: average and maximum
• number of finished jobs

Analysis

Summary of the analysis that have finished on the period:

• identification: Type, Name, Pricing env, Trade filter, type of process
• weight of analysis compared to others
• average response time of a full analysis (all jobs execution + wait time)
• number of executions
• average response time of the pricing preparation
• where it has been executed (remotely, ST, riskserver)

Jobs

Summary of the jobs that have finished on the period :

• identification: Type, Name, Pricing env, Trade filter
• priority of the jobs
• dispatcher configuration where it ran
• weight of jobs compared to others
• average response time of this type of jobs
• number of executions
• average number of pricing executed by jobs
• average and maximum trades computed by jobs

Pricer measure (deactivate by default)

Summary of the pricer measure that have finished on the period :

• identification: Product, Pricer Measure, Process
• weight of pricer compared to others
• average response time of this pricer
• number of executions
• max response time of this pricer

Job detailed execution

This is a single job execution information:

• identification: Type, Name, Pricing env, Trade filter
• priority of the jobs
• dispatcher configuration where it ran
• start time
• duration
• calculator used
• number of trades computed by this job
• number of pricings computed
• number of pricers used
• type of pricing used

Infrastructure

CPU

CPU: The maximum CPU percentage used by DataServer during the selected time period.

(Webview) Over time:

• machine processors count
• percentage of machine total CPU used by the process
• per processor the percentage of the processor CPU used by the process

Memory

• Memory: The maximum Memory allocated to DataServer during the selected time period.

Garbage Collection

GC Activity

• The maximum GC activity as percentage of CPU usage reached by DataServer during the selected time period.

Heap (Webview)

during the selected period, the over time

• Heap bytes in use
• Heap max bytes number
• percentage of Java Heap used (Heap Bytes in use /Heap max bytes number).

Garbage collectors (Webview)

For all Garbage collectors (Old Generation collector and Young Generation collector), the over time

• total number of invocations
• number of invocations per time interval
• ime spent in Garbage collection (per time interval)
• Percentage of time spent in Garbage collection during the last 15 minutes in relation to 15 minutes.
• total time spent in garbage collection

Memory (Webview)

For all memory pools

• Eden space,
• Old generation,
• Survivor space,
• Metaspace

The over time

• Amount of space used by that memory pool
• Current Capacity
• Growth rate over past 1 minute
• Maximum capacity
• Percentage of maximum capactity currently used

Network (deactivate by default)

• Network input bandwith
• Network output bandwith

SocketChanels (Webview)

For all socketChannels, the over time

• number of reading SocketChanels
• number of writing SocketChanels
• input bandwidth
• output bandwidth

For client socketChanels, per server and per port, overtime

• number of writing socketChanel
• output bandwidth

For serverSocketChanels, per port, over time:

• Number of accepts per time interval
• Number of closes per time interval
• Number of writing serverSocketChanels
• Number of reading serverSocketChanels
• Input Bandwidth
• Output Bandwidth

Sockets (Webview)

For clientSocket, per server and per port, over time

• number of opens

Methods Hotspots (Per jvm (dataserver, engines, navigator, risk components…)

Most frequent execution thread stacks during the selected time period by stack cause and client and.The stacks are captured on a regular basis (between 30 sec and 3 min).

The stack cause can be

• STACK for standard stack,
• BLOCKED for blocked thread (usually on a lock),
• BLOCKER for a blocking thread (usually on a lock).

Each thread stack comes with:

• Top Service: The first Calypso calling method in the stack. Excluded packages, classes and methods can be configured.
• Known Stack: the last Calypso calling method in the stack. Excluded packages, classes and methods can be configured.
• Technical Stack: the last method in the stack. Excluded packages, classes and methods can be configured.
• Number of occurrences of the stack at the threads captures.
• The DataServer client, if there is, whose EJB call to DataServer is initiating the stack.
• List of entry/outpoint points detected for this hotspot

Cache Usage (Per jvm (dataserver, engines, navigator, risk components…)

List of Calypso Native Caches used in DataServer and Engines.

Each Cache comes with:

• Reads: The number of times this caches is requested during the selected time period
• Hit Ratio: The percentage of requests that have used cached data during the selected time period
• Population: The max cache number of elements during the selected time period
• Max: Cache max size
• Fill Rate: percentage of Cache population related to Cache max size

Scheduled tasks

The dataserver notification for the execution of scheduled tasks is monitored :

• number of execution by status
• average response time
• total execution time on the period
• task ID / external reference