Meeting with Université de Lausanne
Alex Iribarren
Arnaud Burkhalter
Arne Wiebalck
Daniel Failing
Enrico Bocchi
Grégory Moix
Hamid Hussain-Khan
Jean-François Prévost
Jose Castro Leon
Julien Chimbault
Laurent Borgognon
Luis Fernandez Alvarez
Mikaël Doche
Nicolas Montes
Spyridon Trigazis
Ulrich Schwickerath
Vincent Ducret
Vincent Magnin
Volker Flegel
-
-
08:30
→
09:00
Reception at Science Gateway 30m Science Gateway
Science Gateway
CERN
-
09:00
→
12:00
-
09:00
Welcome & Introductions 5m
-
09:05
Infrastructure 30m
-
OpenStack at CERN 15mSpeakers: Jose Castro Leon (CERN), Luis Fernandez Alvarez (CERN)
-
-
09:35
Storage & HPC 30m
-
HPC at UNIL 15mSpeakers: Jean-François Prévost (UNIL CI/DCSR), Volker Flegel (Université de Lausanne)
-
10:05
Coffee Break 25m
-
10:30
Kubernetes 30m
-
Kubernetes at CERN 15mSpeaker: Spyridon Trigazis (CERN)
-
-
11:00
Network & SDN 30m
-
Network at CERN 15mSpeakers: Daniel Failing (CERN), Vincent Ducret (CERN)
-
11:30
Buffer for Q&A 30m
Who
Question
Answer
JCH
How many VMs do they manage, and how many people are required to manage the OpenStack infrastructure?
JCH
What led to the decision to choose OpenStack at that time?
JCH
Are the VMs billed or is the service free? If they are billed, on what basis? And is this billing automated?
JCH
How is the lifecycle of the VMs managed?
MBA
Use of micro-segmentation (east-west isolation)? If yes, what technology?
MBA
Are VM deployments automated? (cf Aria Operations with automated processes).
MBA
Update cycle (security, updates, new features, etc.)? The level of complexity for carrying out these updates?
MBA
Is their VM architecture solely based on OpenStack or do they operate in a hybrid mode (multiple types of hypervisors)?
MBA
What is their level of satisfaction with the product, support, and OpenStack strategy in the long term?
MBA
What is the operational load (depending on the scope) to manage and maintain the OpenStack infrastructure?
MBA
Do they have a cost estimate for their OpenStack infrastructure (OPEX)?
MBA
Regarding security audits, is it possible to trace information about actions on VMs, hypervisors, etc.?
MDO
Ceph and HPC? How is Ceph used in the context of computation on HPC clusters?
MDO
How are Kubernetes clusters generated on OpenStack?
MDO
Average hardware lifetime?
MDO
How many people are required to manage CERN Ceph clusters?
LBO
Do their OpenStack clusters rely on homogeneous hardware?
LBO
Are they often confronted with solutions that do not support OpenStack as a hypervisor layer?
RFA
How do they manage OpenStack updates?
RFA
How do they manage bare metal deployment?
RFA
Redundant network infrastructure?
-
09:00
-
12:00
→
13:00
-
13:00
→
15:45
Visits: CERN Site
-
13:00
LEIR/LINAC2 45m Low Energy Ion Ring (CERN)
Low Energy Ion Ring
CERN
Speaker: Luis Fernandez Alvarez (CERN) -
14:00
CERN Data Centre 45m Meyrin Data Centre (CERN)
Meyrin Data Centre
CERN
Speaker: Dr Arne Wiebalck (CERN) -
15:00
Antiproton Decelerator (AD) 45m Antimatter Factory (CERN)
Antimatter Factory
CERN
Speaker: Luis Fernandez Alvarez (CERN)
-
13:00
-
08:30
→
09:00