Monitoring Petals CDK-based components

Version 1 by Christophe DENEUX
on Sep 16, 2013 17:39.

compared with
Key
This line was removed.
This word was removed. This word was added.
This line was added.

Changes (53)

View Page History
** message exchange processor object pool metrics,
** message exchange processor thread pool metrics,
** service provider invokcation number per operation and execution status,
** service provider invokcation response times per operation and execution status,
* consumer side:
** no metric.
| {{EnqueuedRequestsCurrent}} | The current number of requests that are enqueued waiting a thread to be processed | GAUGE |

h2. Monitoring the service provider invokcation number per operation and execution status

!service-provider-invokcations-count.png|align=center!

See [Creating a graph using a data query|How to create a graph with Cacti#Creating a graph using a data query] to graph the stats about invokcation number of a service provider:
* the data query definition to put in a file:
{code}
<interface>
<name>Get service provider invokcation number of the BC Soap per operation</name>
<script_path>petals-cli.sh</script_path>
<arg_prepend>-h localhost -n 7700 -u petals -p petals -c -- monitoring -o generic -f service-provider-invokcations-count</arg_prepend>
<arg_index>-- -n petals-bc-soap --list-operations</arg_index>
<arg_query>-- -n petals-bc-soap --query-operations</arg_query>
</filterName>
<succeeded>
<name>Succeeded invokcation counter</name>
<direction>output</direction>
<query_name>SUCCEEDED</query_name>
</succeeded>
<error>
<name>Failed invokcation counter</name>
<direction>output</direction>
<query_name>ERROR</query_name>
</error>
<fault>
<name>Faulty invokcation counter</name>
<direction>output</direction>
<query_name>FAULT</query_name>
</fault>
<pending>
<name>Pending invokcation counter</name>
<direction>output</direction>
<query_name>PENDING</query_name>
* Output fields of the script/command:
|| Name || Description || Type ||
| {{succeeded}} | The succeeded invokcation counter | COUNTER |
| {{error}} | The failed invokcation counter | COUNTER |
| {{fault}} | The faulty invokcation counter | COUNTER |
| {{pending}} | The pending invokcation counter | GAUGE |

h2. Monitoring the service provider invokcation response times per operation and execution status

!service-provider-invokcations-response-times.png|align=center!

See [Creating a graph using a data query|How to create a graph with Cacti#Creating a graph using a data query] to graph the stats about invokcation response times of a service provider:
* the data query definition to put in a file:
{code}
<interface>
<name>Get service provider invokcation response times of the BC Soap per operation</name>
<script_path>petals-cli.sh</script_path>
<arg_prepend>-h localhost -n 7700 -u petals -p petals -c -- monitoring -o generic -f service-provider-invokcations-response-times</arg_prepend>
<arg_index>-- -n petals-bc-soap --list-operations</arg_index>
<arg_query>-- -n petals-bc-soap --query-operations</arg_query>
</filterName>
<succeededMin>
<name>Succeeded invokcation: Min response time</name>
<direction>output</direction>
<query_name>SUCCEEDED_MIN</query_name>
</succeededMin>
<succeededAvg>
<name>Succeeded invokcation: Avg response time</name>
<direction>output</direction>
<query_name>SUCCEEDED_AVG</query_name>
</succeededAvg>
<succeededMax>
<name>Succeeded invokcation: Max response time</name>
<direction>output</direction>
<query_name>SUCCEEDED_MAX</query_name>
</succeededMax>
<succeeded10p>
<name>Succeeded invokcation: 10-percentil response time</name>
<direction>output</direction>
<query_name>SUCCEEDED_10P</query_name>
</succeeded10p>
<succeeded50p>
<name>Succeeded invokcation: 50-percentil response time</name>
<direction>output</direction>
<query_name>SUCCEEDED_50P</query_name>
</succeeded50p>
<succeeded90p>
<name>Succeeded invokcation: 90-percentil response time</name>
<direction>output</direction>
<query_name>SUCCEEDED_90P</query_name>
</succeeded90p>
<pendingMin>
<name>Pending invokcation: Min response time</name>
<direction>output</direction>
<query_name>PENDING_MIN</query_name>
...
<errorMin>
<name>Failed invokcation: Min response time</name>
<direction>output</direction>
<query_name>ERROR_MIN</query_name>
...
<faultMin>
<name>Faulty invokcation: Min response time</name>
<direction>output</direction>
<query_name>FAULT_MIN</query_name>
* Output fields of the script/command:
|| Name || Description || Type ||
| {{succeededMin}} | The min response time of succeeded invokcations | GAUGE |
| {{succeededAvg}} | The average response time of succeeded invokcations | GAUGE |
| {{succeededMax}} | The max response time of succeeded invokcations | GAUGE |
| {{succeeded10p}} | The 10-percentil response time of succeeded invokcations | GAUGE |
| {{succeeded50p}} | The 50-percentil response time of succeeded invokcations | GAUGE |
| {{succeeded90p}} | The 90-percentil response time of succeeded invokcations | GAUGE |
| {{pendingMin}} | The min response time of pending invokcations | GAUGE |
| {{pendingAvg}} | The average response time of pending invokcations | GAUGE |
| {{pendingMax}} | The max response time of pending invokcations | GAUGE |
| {{pending10p}} | The 10-percentil response time of pending invokcations | GAUGE |
| {{pending50p}} | The 50-percentil response time of pending invokcations | GAUGE |
| {{pending90p}} | The 90-percentil response time of pending invokcations | GAUGE |
| {{errorMin}} | The min response time of failed invokcations | GAUGE |
| {{errorAvg}} | The average response time of failed invokcations | GAUGE |
| {{errorMax}} | The max response time of failed invokcations | GAUGE |
| {{error10p}} | The 10-percentil response time of failed invokcations | GAUGE |
| {{error50p}} | The 50-percentil response time of failed invokcations | GAUGE |
| {{error90p}} | The 90-percentil response time of failed invokcations | GAUGE |
| {{faultMin}} | The min response time of faulty invokcations | GAUGE |
| {{faultAvg}} | The average response time of faulty invokcations | GAUGE |
| {{faultMax}} | The max response time of faulty invokcations | GAUGE |
| {{fault10p}} | The 10-percentil response time of faulty invokcations | GAUGE |
| {{fault50p}} | The 50-percentil response time of faulty invokcations | GAUGE |
| {{fault90p}} | The 90-percentil response time of faulty invokcations | GAUGE |