14 | | The first and second evaluation completed found environment and packaging issues, which were resolved by the third evaluation. |
15 | | The third version delivered a version that build without environment issues as documented. Additionally the third version (rev 141) |
16 | | was used to manage both PL and PG resources. |
| 14 | The first and second evaluation completed found environment and packaging issues, which were mostly resolved for the third evaluation. |
| 15 | |
| 16 | The third version delivered a version was built without environment issues. Additionally the third version (rev 141 & 142) |
| 17 | was used to manage both PlanetLab and and ProtGENI resources. Following are findings from these versions: |
| 18 | |
| 19 | * The availability of the pre-built GUSH package simplifies installation and bypasses a tedious process. |
| 20 | * GUSH tool is inconsistent between PL and PG functionality. Users can connect to a PL slice, but must connect to nodes in the PG slices. |
| 21 | * A GUSH restriction for ProtoGENI, is that the ''userid'' must be the same as the ''slicename'' in the GUSH configuration. This is a PlanetLab centric approach and does not match the PG approach where the ''userid'' is used to connect to nodes, not the ''slicename''. |
| 22 | * The new omni feature is not integrated into the gush interface and it is basically the same as running omni from the command line, should this be part of existing GUSH interface, or potentially a new tool that executes slice management actions? |
| 23 | * MyPLC, configuration of required tags for each nodes may be hard to coordinate between MyPLC sites that are part of multiple experiments. |