Risk Register
Likelihood
1: never expected to happen | 2: could happen but very unlikely | 3: could well happen | 4: will probably happen
Impact
1: we can deal with it, no problem | 2: a bit of a hassle but not too bad | 3: can be managed, but with significant effort | 4: crisis
| Category | Owner | Risk Summary | Risk Detail | Risk Likelihood | Risk Impact | Risk Severity | Effect | Mitigation | Comments |
|---|---|---|---|---|---|---|---|---|---|
| Other | CERN/Experiments |
Licensing Limitations |
Licensed software becomes too expensive, or restrictive licensing conditions. |
3 | 1 | 3 | Costs may be unsupportable, or our use cases no longer supported. |
There is very little licensed software still in use in WLCG. The exception is Oracle for services ... |
|
| Software | WLCG |
Loss of support for 3rd party software |
3rd party software may become unsupported. |
3 | 1 | 3 | The loss of key components such as gridftp, OSS tools; would need to replace or ... |
Maintain flexibility and reduce dependencies on single solutions. Components such as gridftp can be ... |
The "grid" is continually evolving, and there are many fewer ... |
| Other | WLCG |
Long-term Reliance on short-term Funding |
Reliance on project-funded activities like EGI risk that support for key ... |
2 | 2 | 4 | Such a situation would require replacement of a key service at short notice or ... |
Ensure that key services have long-term support commmitments within the WLCG collaboration. This is ... |
Such support commitments have been re-iterated several times ... |
| Other | WLCG |
European e-infrastructure diverges from WLCG needs |
EGI or the European Open Science Cloud infrastructures in Europe may potentially ... |
2 | 2 | 4 | A divergence may impose contradictory requirements on participating sites. |
The main mitigation here is continued and strong engagement with the EOSC community and othe science ... |
There has always been a tension between the value of ... |
| Other | WLCG |
GDPR Not Fully Complied with |
The GDPR may not be full comlied with across the full collaboration. |
3 | 2 | 6 | Could have an impact on services and the collaboration of some sites are unable ... |
Due diligence in trying to implement the requirements of the GDPR, and by having clear policy ... |
We have a strong set of policies covering most of the use of ... |
| Software | WLCG |
Experiments' solutions diverge |
Collaboration between experiments breaks down, or solutions for tools and ... |
2 | 3 | 6 | Multiple solutions & duplication of effort too expensive to support in ... |
Ensure common solutions deliver and adapt to majority of use cases. |
Actually there is a very psoitive effort to work on common ... |
| Software | EP/SFT |
Geant4 Does Not Invest in Long Term Performance Improvement |
The Geant4 team (collaboration) does not invest in long term performance ... |
2 | 4 | 8 | Amount of simulation required may not be affordable or achievable with ... |
Invest in a major R&D program to devise strategies for fast MC, as well as better use of compute ... |
This is a long term strategy that must be invested in. Some ... |
| Other | WLCG |
Security Environment changes |
The global scientific security environment could change and break our federation ... |
2 | 4 | 8 | This would be very dispruptive to the operation and require a significant ... |
Ensure our trust networks, policies, awareness are proactively adapting to the changing landscape ... |
While this is a risk, it also also the case that many other ... |
| Software | Experiments |
Experiment Computing Models |
Experiments do not improve computing model performance sufficiently. |
2 | 4 | 8 | Lack of investment of effort to reduce reconstruction time, decrease storage ... |
Experiment working groups on all aspects of computing and storage: nanaAOD, reduce reconstruction ... |
These groups are in place and working, with significant ... |
| Software | HSF/Generators |
Event generators for higher order not optimised |
Event generators for higher orders are not optimised |
2 | 4 | 8 | Event generation will be expensive in CPU time and become a significant fraction ... |
Workshops organised via the HSF with the generator community. Must keep pushig this as a priority |
Generating events at higher orders will become more ... |
| Software | EP/SFT |
ROOT Does Not Invest in & Prioritise I/O performance and data formats |
The ROOT team may not give enough priority to ensuring ROOT data structures and ... |
2 | 4 | 8 | I/O performance coould become a significant bottleneck in data performance ... |
Start R&D program on I/O and data management in ROOT, prioritse above other functions; has major ... |
This is an area where improvements will benefit all ... |
| Technology | WLCG |
Change of HDD Technology is expensive or difficult to use |
New technology for HDD could be difficult to use (i.e. no longer random access) ... |
2 | 4 | 8 | The overall effect would be an increase in storage costs (for disk). |
The future data lake model is part of a mitigation for this scenario. Focussing data storage on few ... |
The increased use of tape to mitigate disk costs, however ... |
| Technology | WLCG |
Tape market shrinks/fails |
The market for tape shrinks or fails completely, due to loss of revenues ... |
2 | 4 | 8 | Major rethink of value of data vs re-run of experiment, Need new data models ... |
The future data lake model is part of a mitigation for this scenario. Focussing data storage on few ... |
We have already seen Oracle pull out of Enterprise tape ... |
| Funding | CERN |
Flexibility of CERN project budget lost |
The project nature of the CERN LCG budget is important, the risk is this becomes ... |
2 | 4 | 8 | Cannot adapt spending to often changing and unpredictable LHC parameters and ... |
Work with RPC to keep the project nature of the budget. |
|
| Software | WLCG/HSF |
Lack of Investment in Software Skills |
The community and funding agencies fail to re-invest in advanced software skills ... |
3 | 3 | 9 | Inability to improve performance of the experiments' cores software, or common ... |
Training, education, hackathon programmes with HSF; projects such as IRIS-HEP, & other initiatives. ... |
This is a problem faced in many sciences that depend on ... |
| Technology | WLCG/GEANT/NRENs |
Inadequate WAN; or competition with other sciences |
Wide Area Networking bandwidth does not provide the bandwidth we need to support ... |
3 | 3 | 9 | Lack of sufficient bandwidth to support data delivery models. |
We must work with the networking community to exploit technology innovations to permit traffic ... |
Until now LHC has been given as much network capacity as it ... |
| Technology | WLCG |
Requirement of some FA's to provide HPC or GPU as part of the pledge may lead to resources matched to requirements, or could mean that some resources can only be used for some workflows. |
Some architectures are not well adapted to the production workflows of the ... |
3 | 3 | 9 | Inefficient usage; some resources may only support certain workflows; loss of ... |
As with other technology related risks, the mitigation of to ensure that the experiments' core ... |
Similar conclusions to other technology and funding related ... |
| Technology | WLCG |
Offered resources not matched to requirements (e.g. HPC, GPS, etc) |
Requirement of some FA's to provide HPC or GPU as part of the pledge may lead to ... |
3 | 3 | 9 | Some architectures are not well adapted to the production workflows of the ... |
As with other technology related risks, the mitigation of to ensure that the experiments' core ... |
Similar conclusions to other technology and funding related ... |
| Funding | Experiments |
Physics Requests for Resources not Affordable |
The funding agencies are unable to provide the funding to purchase the required ... |
3 | 3 | 9 | Cannot provide required resources to experiments; Not all physics programmes can ... |
Prioritise the physics programme of the experiments. |
If the overall level of funding is significantly lower than ... |
| Other | WLCG DOMA |
Lack of Operational Support at Sites |
Lack of adequate operational support (staff) at sites. |
3 | 3 | 9 | Would lead to unreliable services and unresponsive response to operational ... |
The majority of key services are now already run either at CERN or a few other Tier 1 stes. The ... |
The other factor here is that many sites today also need to ... |
| Funding | CERN |
CERN Tier 0 budget insufficient. |
The Tier 0 budget (MTP) is not adequate to fulfill the CERN commitment to LHC ... |
3 | 4 | 12 | Cannot fulfill CERN commitments; knock-on effect to other Funding Agencies who ... |
Reduce scope of the Tier 0 responsibility. |
The Tier 0 budget for Run 3 is less than anticipated needs ... |
| Funding | Funding Agencies |
WLCG Funding Agencies lack funding |
The funding agencies are unable to provide the funding to purchase the required ... |
4 | 3 | 12 | Cannot provide required resources to experiments; Not all physics programmes can ... |
Prioritise the physics programme of the experiments. |
If the overall level of funding is significantly lower than ... |
| Funding | CERN |
No PCC for Run 4 |
There is no new Computer Centre (PCC) available in time for Run 4 |
3 | 4 | 12 | CERN will be unable to deliver the Tier 0 commmitments. There will be no ... |
Significantly reduce the scope of the Tier 0 commitment. |
Similar consequences to Risk 1. In addition possible ... |
| Technology | WLCG |
Technology Evolution Insufficient |
The evolution of the costs of compute, storage, and other key technologies ... |
4 | 4 | 16 | We are unable to provide the needed level of resources with constrained budgets. |
Ensure the core software of the experiments is as efficient as possible, also the experiments may ... |
The evolution of "Moore's Law" over the past several decades ... |