Nagios requirements

From EGIWiki
Jump to: navigation, search
Main EGI.eu operations services Support Documentation Tools Activities Performance Technology Catch-all Services Resource Allocation Security


Submitting a Nagios test requirement

Operations Tool Advisory Group (OTAG) members can submit Nagios probe requirements on behalf of their NGI and their resource centres. Requirements are collected on the requirements queue hosted on the EGI RT system. These are regularly collected on a quarterly basis, but requirements can also be asynchronously submitted when needed.

Who is responsible of processing Nagios test tickets? Nagios test requirements are discussed and prioritized in the framework of OTAG at least on a quarterly basis, depending on the urgency of the requirement.

Who is authorized to open a ticket?

NGI/EIRO Operations Centre Managers that are members of the noc-managers SSO group and SA1 task leaders.

Opening a ticket on RT

Go to RT, select the requirements queue (top right-hand corner) from the drop-down menu, and select New ticket.

Setting a Nagios probe requirement ticket

This section provides information on how to properly set the ticket metadata:

  • Owner: "Daniele Cesini" from the drop-down menu (this as an indirect way to indicate that the OTAG is the owner of the ticket)
  • Category (leve 1): "Unified Middleware Distribution" from the drop-down menu
  • Category (level 2): "Monitoring" from the drop-down menu
  • Requestor (level 1): "NGI or EIRO" from the drop-down menu
  • Requetor (level 2): select the name of your NGI, or "Other" in case your NGI does not appear in the drop down menu
  • Reqiestor (level 3): "no value" from the drop-down menu
  • Technology Tag: select the specific middleware distribution if relevant, or specify "no value" if the requirement is applicable to all stacks
  • Custom Tag: "new nagios probe1/modify nagios probe2/request nagios probe3"

1In case of a Category 1 requirement

2In case of a Category 2 requirement

3In case of a Category 3 requirement

  • Non Functional Tag: select what is applicable from the drop-down menu, depending on the type of requirement

Priority

Priority can be set at the moment of the creation of the ticket. In order to do so, use the Impact field. The priority levels that can be selected in the impact field are in a range of 4 levels, from low to high as follows:

  • Don't know
  • Nice to have - slight improvement
  • Useful – will be of some benefit to this community
  • Important with a measurable benefit to this community
  • Essential to the effectiveness of this community/the wider this community

Nagios test requirement categories

Category 1

This category includes requirements for probes developed by NGI to be integrated into SAM.

A ticket can be submitted in this category if the NGI/EIRO wishes to share a probe internally developed that may be potentially applicable and useful to the whole EGI.

For each probe in this category, the requesting NGI/EIRO should specify the following information in the body of the ticket:

  • Grid service that probe tests: specify which feature of the service is tested by the probe (e.g. CPU load)
  • Brief description of test: provide a high-level description of the test including the conditions that generate a WARNING and CRITICAL result.
  • Developer contact information
  • Link to existing documentation (if any)

Category 2

This category is applicable to tickets that are submitted to request changes to existing probes, in the case an NGI/EIRO needs to require a change to a test already distributed.

For each probe in this category, the requesting NGI/EIRO should specify the following information in the body of the ticket:

  • Test name
  • Brief description of changes requested: please provide here information on what should change, and provide information on a use case that justifies this change.
  • Is NGI prepared to develop changes?

Category 3

A ticket can be submitted to request that a test is developed. Tests that are relevant are those applicable to the whole NGI. Tests that are requested for regional deployment and just address local requirements are out of scope.

For each probe in the body of the ticket provide the following information:

  • Grid service that should be tested: provide information on
    • which metric the probe should return
    • other specific probes currently existing for this service.
  • Write a brief description of the probe, this including a use case.
  • Is NGI prepared to develop probes?