The MAAP/HECC project may require the construction of infrastructure or specialized systems to support various different processing capabilities. These capabilities may require the deployment of physical systems, or involve contracting for services or capabilities external to HECC.
The list of potential specialized needs that may be required include:
1) provide persistent NAT services for compute nodes.
2) provide front end capability to MAAP/HEC for R2
3) staging of data, both for input and output
4) caching of data from several different sources of large cloud based data sets.
5) the support of different types of batch oriented processing that require tracking progress, but require very little 'compute' resources as might be used for staging data.
6) specialized systems or network configurations to facilitate higher bandwidth WAN connectivity
7) acquisition of cloud specific services to support efficient operation (e.g. AWS direct connect)
8) identify storage/filesystem that can be used for R2/R3 runs
Definition of Done:
[x] Deliverable 1 - Mechanism for NAT services on compute nodes
[x] Deliverable 2 - Provide front end capable of supporting R2
[ ] Deliverable 3 - Deploy any additional border facing system - HFE2 - for R3.
[ ] Deliverable 4 - Upgrade or replace border facing systems with 100gb to match WAN capabilities.
[ ] Deliverable 5 - Assess need for PBS to efficiently allocate multiple jobs to a single or few border systems
[ ] Deliverable 6 - Provide capability to access AWS efficiently delineating steps or requirements for AWS direct connect at 10 and 100gb.
The MAAP/HECC project may require the construction of infrastructure or specialized systems to support various different processing capabilities. These capabilities may require the deployment of physical systems, or involve contracting for services or capabilities external to HECC.
The list of potential specialized needs that may be required include:
1) provide persistent NAT services for compute nodes. 2) provide front end capability to MAAP/HEC for R2 3) staging of data, both for input and output 4) caching of data from several different sources of large cloud based data sets.
5) the support of different types of batch oriented processing that require tracking progress, but require very little 'compute' resources as might be used for staging data. 6) specialized systems or network configurations to facilitate higher bandwidth WAN connectivity 7) acquisition of cloud specific services to support efficient operation (e.g. AWS direct connect) 8) identify storage/filesystem that can be used for R2/R3 runs
Definition of Done: