Regional operations

From GridPP Wiki
Jump to: navigation, search

Moving towards regional operations

UKI is currently scheduled to move away from the central/global COD model of operations to a regional model. This move is planned during the EGEEIII project to reduce central operational commitments and help countries move towards their National Grid Initiative mode (expected as part of the transition to the EGI world). UKI moves to a regional basis from 15th June 2009. This page lays out the work areas and what will be or has been put in place.

R-COD arrangements

People: In the COD model Tier-1 staff led UKI engagement with the on-duty tasks. Leading up to June the UKI EGEE coordinators (one per Tier-2 and who work alongside the GridPP Tier-2 Technical Coordinators) and colleagues from Grid-Ireland have been taking over the task and they will perform the regional on-duty work from June 15th.

Shifts: The current plan is to have two people on shift in each weekly period.

Action: We need to create a page for the shift list.

Action: We need to agree the people on shift for the next 6-12 months


Tools: For regional operations there is a regional version of the COD dashboard. This will be used by those on-duty to raise and track issues. For those with sufficient access priveledge the R-COD dashboard is here: https://cic.gridops.org/index.php?section=roc&page=dashboard. Areas of the dashboard can be summarised as follows:

Tickets

Alarms

Status dashboard

SAM submission tool

Downtimes

Handover

Regional view (to see status in other regions)

Metrics (based on tickets opened, response times etc.)

Preferences (allows selection of viewing preferences)


Action: Agree and document handover procedure

Communications

1) A UKI grid-ops list has been created under JISCMAIL. This list is used by the on-duty staff to exchange information.

Action: Decide on jabber/skype meeting room for on-duty communications. Use of weblog for recording progress?


First line support

In the regional model first line support for problems encountered in the region is a regional responsibility. We already use TB-SUPPORT to seek support within the sysadmin community, but this needs to extend to user and application problems.

Action: Agree the structure to be implemented for:

People - which list to use?

Shifts - at least one person to lead and ensure follow-up each week

Tools & share of responsibilities - GGUS tickets based?

Communication - another list?

C-COD - our representatives

Knowledge sharing - a weblog?

Short term problems

Regional helpdesk


Checklist

At least one FTE to do this work

Provide mailing list for R-COD operators (as a contact point) ‏ Send a link to the GGUS for regional knowledge base (if it exists) ‏ Operators - 1st line support, R-COD, C-COD

Have a valid certificate

Read the operations manual

1st line supporters must request “regional first line supporter” role in the GOCDB

R-COD must request “Regional Staff” role in the GOCDB and be on the relevant R-COD mailing list.

C-COD must request “C-COD Staff” role in the GOCDB and sign up to the C-COD mailing list.

Choose correct preferences in the CIC regional dashboard