PROC04 Quality verification of monthly availability and reliability statistics

From EGIWiki
(Redirected from PROC04)
Jump to: navigation, search
Main EGI.eu operations services Support Documentation Tools Activities Performance Technology Catch-all Services Resource Allocation Security


Documentation menu: Home Manuals Procedures Training Other Contact For: VO managers Administrators


Contents


Title Quality verification of monthly availability and reliability statistics
Document link https://wiki.egi.eu/wiki/PROC04
Last modified 16th August 2018
Policy Group Acronym OMB
Policy Group Name Operations Management Board
Contact Group operations@egi.eu
Document Status Approved
Approved Date 30 October 2012
Procedure Statement Instructions RODs and Operations Centres on how to handle justification for poor monthly performance
Owner Alessandro Paolini


Overview

The document describes the process of how to handle justification for poor monthly performance.

Links to all monthly statistics are provided on a regular basis at Availability and reliability monthly statistics page.

Definitions

Please refer to the EGI Glossary for the definitions of the terms used in this procedure.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", “MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.

Process of handling RC Availability and Reliability

Availability alarms are raised on the ROD Dashboard and are thought to be a warning for NGI informing about poor performance of site within the last 30 days.

Entities involved in the procedure

Regional Operator on Duty (ROD): team provided by NGIs and responsible for handling RC availability and reliability alarms through Dashboard in Operations Portal.

Operations: team provided by EGI.eu and responsible for handling of underperforming sites which were below the target for 3 consecutive months.

NGI manager: person who suspend the underperforming site or provide site justification.

Steps

When an alarm is raised, it means that the Availability metric has dropped below the threshold of 80% for the last 30 days period.

Handling alarms:

Step# Responsible Action
1 ROD Creates a ticket through the dashboard notifying site administrator that the Availability metric has dropped below the threshold of 80% for the last 30 days period.

The expiration date should be set to not later then same date next month minus 1 day.

2 ROD Escalation of the ticket will vary between NGIs. NGIs have freedom to decide if they want to apply any escalation procedure or treat availability tickets just as an notification for site administrators.
3 ROD
  • Ticket can be closed as soon as the alarm goes into OK status (however it is recommended to make sure it is a couple percent above the threshold before closing it. e.g. 85%).
  • If the problem continues for over 30 days the ticket should be closed otherwise ticket will appear on Operations dashboard and affect ROD performance index.

Handling of underperforming sites (below the target for 3 consecutive months):

Step# Responsible Action
1 Operations Creates a GGUS ticket for each underperforming site. Ticket template.
2 NGI operations manager Within 10 working days NGI operations manager can suspend the site or ask to not suspend the site by providing adequate explanation
3 Operations Send a direct email to NGI and site contact email (in GOC DB) with deadline 2 days for comments
4 Operations
  • In the case of no NGI intervention, the site is suspended in GOC DB.
  • In the case of NGI intervention:
    • non suspension will occur if the Operations team agree on the reasoning provided by the NGI (the Chief Operations Officer may be involved)
    • if availability shows no improvement Operations can suspend the site

Recomputation procedure

In case of doubts about the validity of Availability/Reliability reports, a RC/NGI can request recomputations according to the procedure defined at PROC10

Known issues and recommendations to NGIs

  1. Newly certified sites will get inaccurate Availability/Reliability figures for the month they were certified and all months before that. ARGO Computation Engine takes into account the Certification status of the site in GOCDB in order to decide if metrics should be calculated for the site. Because the Certification status history is not currently available in the operations tools, until a solution is implemented NGIs should check if they have sites affected by this issue and report it as explanation. More information at [1] and [2].
  2. Recalculation - The calculations performed by ARGO always take into account the information system status and GOCDB information at the time the calculation is performed, and not that of a certain checkpoint in the past. The implication of this is that any complete recalculation has the risk of altering the results for sites that had correct numbers in the first place. Thus until a solution is found, complete recalculations are avoided whenever possible, and errors are fixed on per site basis for those that have lower number than they should.
  3. Weighted availability is calculated by multiplying the number of logical CPUs a site published with the published HEPSPEC value in BDII. It is important that these numbers are correct, if HEPSPEC for a site is too high or too low (for example in case of mistake) the overall NGI weighted availability will be affected.

Revision history

Version Authors Date Comments
M. Krakowian 19 August 2014 Change contact group -> Operations support
Alessandro Paolini 2016-06-08 Change contact group -> Operations
Alessandro Paolini 2018-08-16 A/R of NGI Core services no more handled, deleted from procedure; updated some links



Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox
Print/export