Minutes for Thursday

 

Network meeting

- We'll get more bandwidth

- We want 10x increase to CERN

- We want dedicated links rather than shared to CERN

- We want 10x increase to sites

- Finland has some connection issues which causes drops in TCP sometimes

 

Site Expectations

- No one seems to be properly monitoring sites for dcache nodes

- All sites need to be better about monitoring their dcache pools and not just depend on NeiC

- We'll create a wiki document to clarify monitoring requirements for sites for dcache pools

- Discussing setting full expectations of how to monitor sites.

   - dCache

   - ARC

   - hammercloud

   - etc

- Wiki pages need to be maintained better meaning less duplication and more use of links

- People want to have a more clear list of who is responsible for what

- Jens is making progress on cleaning up the wiki for how to install dcache pools

- Rename monitoring section of dcache wiki to metrics

- As always, we'll talk more about wiki cleanup at the next meeting

 

Monitoring

- grafana.ndgf.org