3: Excellent Services and Operations by

Need a definition of excellence

2023-12-04 16:26:15 UTC

Limited funding to cover all the priorities

2023-12-04 16:26:26 UTC

Priorities - hard to actually rank

2023-12-04 16:30:17 UTC

3B - important to ensure longevity of services that people rely on

3C - always important

3A (though similar to 3E?)

3E (similar to 3A?) - some doubt as to feasibility of federating everything

Need actual objectives mapped to these priorities

2023-12-04 16:31:52 UTC

The priorities are so generic that it's hard to find anything missing, for example.

The objectives for these priorities

2023-12-04 16:32:19 UTC

Engage with existing UKRI etc. areas

2023-12-04 16:33:34 UTC

e.g. sustainability

Priorities mean diff things to diff communities/groups

2023-12-04 16:34:43 UTC

Lack of effort in Ops

2023-12-04 16:35:26 UTC

Lack of control

2023-12-04 16:35:58 UTC

project management, federated environment

Hard to say due to generic priorities

2023-12-04 16:36:18 UTC

Priorities (all v. important)

2023-12-04 16:41:10 UTC

3C - Security is always a top priority

3D - better meeting community needs through service evolution

3E - resilience and flexibility

3B - Long term running is key

3A - Creating accessible infra is critical to uptake

Operational Priorities

2023-12-04 16:41:35 UTC

3B: the IRIS service needs to be available in the longer term (at least 5 yr planning horizons).

3C: IRIS needs to offer a secure infrastructure

3E: IRIS needs to be able to accommodate a wide range of end user science projects/ communities.

3D:

3A: it depends on the 'services'

Longer term planning:

2023-12-04 16:41:53 UTC

Are we talking about people here ? Because for excellent operations we need experienced people in place.

Operations

2023-12-04 16:44:05 UTC

If IRIS becomes an operational infrastructure then there needs to be a focus on operational support. So a ticketing system, helpdesk, monitoring usage of the resources. This also involves connecting with the support services available at the resource providers

2023-12-04 16:47:43 UTC

I don't quite understand 3B (the English doesn't work) because it conflates two things: longer term planning... thus making the services more sustainable (in the funding sense); but then it also talks about improving service efficiency, which is a separate issue surely? What services are inefficient and why is that?

Hosting and web presence facilities for projects

2023-12-04 17:07:53 UTC

e.g. it took a long time for ExCALIBUR, for H&ES, for ExCALIBUR software projects to develop their own web presence (and some still haven't). Having a central facility that becomes known for allowing this would be great - publish webpages, publish internal confluence-type pages, etc.

B, C, A, E, D

2023-12-04 17:09:44 UTC

B is essential - so that communities can plan and understand they will have access to reliable and predictable services. Security is second as securing the services is very important, if they're compromised they are unusable.

Federated infra services created a lot of discussion about this was - useful for some, but not for all. It may be it can be implemented but the netork performance required has so far been unattainable.

D - sounds like it limits the scope, providers shouldn't tell the users what it is they want.

Operational Scheduling

2023-12-04 17:10:36 UTC

In managing big project use of IRIS, in terms of DC and AC processing - the ability to schedule across IRIS will become increasingly important. Many projects know when they will need significant access to processing resources, and when they will have less need. The ability to schedule, especially with a year ahead horizon, would be very useful.

Provision of reliable long-term funding plan

2023-12-04 17:13:38 UTC