| Name | Bulk data multiplexing |
|---|---|
| Category | System Test |
| Subject | A/V streaming |
| Subsystems concerned | ACS, Correlator, Archive, TelCal, QuickLook |
| Test design | HLA/Archive |
| Test execution | ITS/HLA |
| Description | The main concern of the BulkStore is the required throughput between the Correlator and the final location of the data on NGAS. The Archive subsystem does carry out tests on the individual components of the archive part of this chain and there are also tests of the plain A/V streaming. More advanced testing is necessary at some point to simulate the required multiplexing of the data from the Correlator to the Archive, TelCal and QuickLook subsystems. This involves four subsystems and the archive is the least critical one in terms of timing requirements. |
| Goal | Verify and quantify the ability of the adopted A/V streaming solution to deliver the required data rate simultanously to Archive, TelCal, QuickLook. |
| Number | Prerequisite | Subsystems involved |
|---|---|---|
| 1 | Stable and agreed A/V streaming library | ACS, Correlator, Archive |
| 2 | Sender and receiver implementations for all subsystems | Correlator, Archive, TelCal, QuickLook |
| 3 | Deployment design | HLA, ITS |
| 4 | Adequate hardware setup | ITS |
| 5 | Definition of the required throughput for the setup in (3) | HLA |
| Name | Bulk data Load Balancing |
|---|---|
| Category | System Test |
| Subject | A/V streaming, NGAS storage |
| Subsystems concerned | Archive, Correlator |
| Test design | Archive |
| Test execution | Archive |
| Description | This test |
| Goal | Verify and quantify the ability NGAS using the adopted A/V streaming solution to store data with the expected maximum throughput for a given deployment and scalability to 60 MB/s and with the required stability (min. 24 hours) |
| Number | Prerequisite | Subsystems involved |
|---|---|---|
| 1 | Implementation of A/V streaming chain with BulkSender BulkDistributor and BulkReceiver | ACS, Correlator, Archive |
| 2 | Implementation of load balancing BulkReceiver | Archive |
| 3 | Deployment design | Archive, ACS, ITS |
| 4 | Implementation of deployment on adequate hardware setup | ITS, Archive |
| 5 | Definition of the required throughput for the given setup in (3) | HLA |
| Name | Monitor Ingest Rate |
|---|---|
| Category | Archive Test |
| Subject | A/V streaming, Transactions |
| Subsystems concerned | Control, ACS, Archive |
| Test design | Archive |
| Test execution | Archive |
| Description | |
| Goal | Verify and quantify the ability of the MonitorStore to handle the required data rate. |
| Number | Prerequisite | Subsystems Involved |
|---|---|---|
| 1 | Stable and agreed A/V streaming library | ACS |
| 2 | Stable and agreed monitor format, including definition of key structure for monitor devices. | Control |
| 3 | Detailed data rate numbers, including expected number of devices and monitor points, data types and formats. | Control |
| 4 | Definition of the required rate using (3) | Archive, Control |
| Name | Test Monitor Query |
|---|---|
| Category | Archive Test |
| Subject | Query usability and performance |
| Subsystems concerned | Control, Archive |
| Test design | Archive |
| Test execution | Archive, External |
| Description | |
| Goal | Verify and quantify the usability and performance of typical queries against the MonitorStore. |
| Number | Prerequisite | Subsystems Involved |
|---|---|---|
| 1 | Stable and agreed monitor format, including definition of key structure for monitor devices. | Control |
| 2 | Definition of typical queries, including typical time ranges and transaction rates. | Control |
| 3 | Definition of the required rate using (2) | Archive, Control |
| Name | APDM Ingest |
|---|---|
| Category | Archive Test |
| Subject | APDM ingest performance |
| Subsystems concerned | OT, Archive, SSR |
| Test design | OT, Archive |
| Test execution | OT, Archive, External |
| Description | Test covers the data structures as defined by ObsPrep. Since this is a hierarchical, recursive structure the actual performance is best being verified using realistic data. In preparation of the APDM query test, it would be benefitial to get complete real-world projects with all parameters, SBs and scripts included. |
| Goal | Verify and quantify the ingest performance of APDM data structures. |
| Notes | This test can be done in a staged approach, first using dummy incomplete APDM, and then upgrading to real projects, like the DSRPs. |
| Number | Prerequisite | Subsystems Involved |
|---|---|---|
| 1 | Stable and agreed APDM format. | ObsPrep |
| 2 | Definition of the required rate using the DSRP numbers (?) | Offline, SSR |
| Name | APDM Query |
|---|---|
| Category | System Test |
| Subject | APDM query usability and performance |
| Subsystems concerned | Archive, and all others, depending on complexity. |
| Test design | HLA |
| Test execution | ITS/HLA |
| Description | Test covers the retrieval of the data structures as defined by ObsPrep. Since this is a hierarchical, recursive structure the actual performance is best being verified using realistic data. The level of complexity can be increased step-wise, i.e. in the first stage just (dummy) OT APDM, second stage with referenced (dummy) exec records (Control) and third stage with referenced ASDM and bulk entities. |
| Goal | Verify and quantify the ingest performance of APDM data structures. |
| Notes | Query optimisation in the archive will only be started once the APDM is relatively stable and complete. Real tests can only be started once there is a reasonable amount of realistic projects available. |
| Number | Prerequisite | Subsystems Invloved |
|---|---|---|
| 1 | Stable and agreed APDM format. | ObsPrep |
| 2 | Definition of the queries based on TheQueryPage and the ASA Use Cases | Archive |
| 3 | Availability of various realistic and complete programs | SSR using OT (?) |
| 4 | Availability of exec records | Control |
| 5 | Availability of ASDM entities | Capture |
| 6 | Availability of bulk entities | Correlator |
| Name | ASDM Ingest |
|---|---|
| Category | Archive/System Test |
| Subject | ASDM ingest performance |
| Subsystems concerned | Control(Capture), Archive |
| Test design | Archive, Control |
| Test execution | Archive |
| Description | Test covers the science data structures as defined by Offline and HLA. Since this is a rather complex structure the actual performance is best being verified using realistic data. In preparation of the ASDM query test, it would be benefitial to get complete real-world projects catured with all parameters included. |
| Goal | Verify and quantify the ingest performance of ASDM data structures. |
| Notes | This test can be done in a staged approach, first using dummy incomplete ASDM, and then upgrading to real projects, like the DSRPs. |
| Number | Prerequisite | Subsystems Involved |
|---|---|---|
| 1 | Stable and agreed ASDM format. | Offline |
| 2 | Evaluation of the required rate using the DSRPs (?) | SSR |
| Name | ASDM Query |
|---|---|
| Category | System Test |
| Subject | ASDM query usability and performance |
| Subsystems concerned | Archive, and all others, depending on complexity. |
| Test design | HLA |
| Test execution | ITS/HLA |
| Description | Test covers the retrieval of the data structures as defined by ObsPrep. Since this is a hierarchical, recursive structure the actual performance is best being verified using realistic data. The level of complexity can be increased step-wise, i.e. in the first stage just (dummy) Capture ASDM, second stage with reference (dummy) bulk entities and third stage with simulated or real bulk entities. |
| Goal | Verify and quantify the query functionality against ASDM data structures. |
| Notes | Query optimisation in the archive will only be started once the ASDM is relatively stable and complete. Real tests can only be started once there is a reasonable amount of realistic projects available. |
| Number | Prerequisite | Subsystems Invloved |
|---|---|---|
| 1 | Stable and agreed ASDM format. | Offline |
| 2 | Definition of the queries based on TheQueryPage and the ASA Use Cases | Archive |
| 3 | Availability of various realistic and complete captured program | HLA |
| 4 | Availability of exec records | Control |
| 5 | Availability of APDM entities | ObsPrep, SSR (DSRPSs)) |
| 6 | Availability of bulk entities | Correlator |
| Name | ASA Query |
|---|---|
| Category | Archive Test |
| Subject | ASA query usability and performance |
| Subsystems concerned | Archive, SSR |
| Test design | Archive |
| Test execution | Archive, SSR, External |
| Description | Test covers the retrieval of the data structures as defined by ASA. Since this is complex structure the actual performance and usability is best being verified using realistic data. The level of complexity can be increased according to the priorities defined in the ASA use-cases. |
| Goal | Verify and quantify the query usability and performance of ASA data structures. |
| Notes | Query optimisation in the archive will only be started once the ASA is relatively stable and complete. Real tests can only be started once there is a reasonable amount of complete realistic projects available. The usability could be evaluated before data is actually available. |
| Number | Prerequisite | Subsystems Invloved |
|---|---|---|
| 1 | Stable and agreed ASA data model. | Archive |
| 2 | Definition of the queries based on the ASA Use Cases | Archive |
| 3 | Availability of various realistic and complete programs | SSR using OT? |
| 4 | Availability of exec records | Control |
| 5 | Availability of ASDM entities | Capture |
| 6 | Availability of bulk entities | Correlator |
| Name | Archive Concurrency |
|---|---|
| Category | System Test |
| Subject | Archive stress test with concurrent access. |
| Subsystems concerned | Archive, HLA, ITS |
| Test design | HLA |
| Test execution | ITS/HLA |
| Description | Test covers the ingest and retrieval of any kind of data in a concurrent way. Concurrent performance is expensive, because it needs first good planning and then optimal deployment of the different parts and clients. |
| Goal | Verify and quantify the archive performance under stress. |
| Notes | Concurrent access to the archive has to be planned, because this will cause the highest impact on the raw performance figures from all the other individual tests. |
| Number | Prerequisite | Subsystems Invloved |
|---|---|---|
| 1 | Analysis of the ALMA data flow model, including rate and timing analysis. | HLA |
| 2 | Availability of either real or simulated clients, which implement the model from (1) | ITS |
| 3 | Realistic hard- and software deployment | ITS |