Project

General

Profile

Edit Actions
Copy link

Feature #20094

open

Alerting and Monitoring facilities for ISTI Open Portal

Added by Leonardo Candela about 4 years ago. Updated over 3 years ago.

Status:
New
Priority:
High
Assignee:
Andrea Dell'Amico
Target version:
ISTI Open Portal Tasks
Start date:
Nov 09, 2020
Due date:
% Done:

0%

Estimated time:
Quote

Description

We should design and deploy a number of monitoring approaches aiming at alerting the service managers about any malfunction.

I see at least two families:

I do ignore what we have in place right now, certainly we do not have any alert raised when the primary data source (not operated by ISTI) is not responding ... or I'm not receiving any notification.

Add

Subtasks

Add

Related issues

Related to ISTI Open Access - Support #22008: Cater for OpenPortal for ISPCClosedLeonardo CandelaSep 15, 2021

Delete relationActions
QuoteActions
Copy link
 #1

Updated by Andrea Dell'Amico about 4 years ago

We do not have anything at the moment, the monitoring system for ISTI is still a work in progress.

Actions
Copy link
 #2

Updated by Leonardo Candela over 3 years ago

  • Priority changed from Normal to High
Actions
Copy link
 #3

Updated by Leonardo Candela over 3 years ago

QuoteActions
Copy link
 #4

Updated by Leonardo Candela over 3 years ago

We recently faced with an issue on People side, i.e. we have been provided with a smaller number of items. Of course this is an application oriented "malfunction" and we should identify an approach for alerting the right person(s).

In essence, there are at least two typologies of monitoring and alerting we should have:

  • alerts on machines and services "availability", "loads";
  • alerts on services business logic malfunctions;
QuoteActions
Copy link
 #5

Updated by Andrea Dell'Amico over 3 years ago

Leonardo Candela wrote in #note-4:

We recently faced with an issue on People side, i.e. we have been provided with a smaller number of items. Of course this is an application oriented "malfunction" and we should identify an approach for alerting the right person(s).

In essence, there are at least two typologies of monitoring and alerting we should have:

  • alerts on services business logic malfunctions;

We need help from the service to handle this one. I guess a set of prometheus friendly metrics should suffice.

QuoteActions
Copy link
 #6

Updated by Leonardo Candela over 3 years ago

I'm fine with Prometheus and Grafana and willing to play with them:

  • Do we have any dashboard for Open Portal? What shall I do to access it?
  • Is there any "application context" defined for it?
  • Is @michele.artini@isti.cnr.it aware on how to produce "prometheus friendly metrics"?
QuoteActions
Copy link
 #7

Updated by Andrea Dell'Amico over 3 years ago

Leonardo Candela wrote in #note-6:

I'm fine with Prometheus and Grafana and willing to play with them:

  • Do we have any dashboard for Open Portal? What shall I do to access it?

Not yet

  • Is there any "application context" defined for it?
  • Is @michele.artini@isti.cnr.it aware on how to produce "prometheus friendly metrics"?

I don't know. I also don't know if there's something specific to Joomla, but there are a lot of php-fpm exporters: https://github.com/Lusitaniae/phpfpm_exporter, https://bidhankhatri.com.np/monitoring/monitoring-php-fpm-pools-through-prometheus-and-grafana-on-centos-6-and-7/, https://github.com/hipages/php-fpm_exporter

Edit Actions
Copy link

Also available in: Atom PDF