Telescope Manager

TelescopeManager_blue

As indicated in the last edition of the SKA eNews, following the Telescope Manager (TM) CDR closure by the end of July 2018, the Bridging Phase for the TM was officially started from the start of August 2018. The TM consortium ceased to exist post July 2018 and in the post-consortium era, the SKA Office needed the individual countries and institutes commit for the Bridging Phase effort. Responding positively, India committed seven FTEs for the bridging phase and continued contributing. This was important in order to preserve the skills and competencies developed during the Design Phase.

The SKAO has adopted Scaled Agile Framework (SAFe) methodology during the Bridging Phase for its software development activities and the team from India is working as one of the Agile Team. As part of the process, the important System CDR related activities and risks coming out of TM CDR are managed and prioritised at the SAFe Programme level by the Product Management Team (people from the SKAO) and then assigned to and taken up by the Agile Teams in bi-weekly Sprints. By the end of November, the Indian Agile Team completed six such Sprints.

The summary of the major work taken up by the Indian Agile Team in the last six Sprints since the TM CDR closure, key events and accomplishments are as follows:

Development of TMC Prototype – Stage I:

We had been tasked to develop and demo Telescope Monitor and Control (TMC) evolutionary prototype at the time of Programme Increment (PI) Planning meeting scheduled during the first week of December. The scope and objective of the prototype is as follows

  • Adoption of TMC Architecture/Design
  • Utilization of SKA LMC Base classes
  • Implementation of TM-Dish interface
  • Use of TANGO Platform tools – Elettra, LogViewer, Pogo and Jive

The team was able to develop the prototype in a short time and demoed during the PI Planning meeting and so far the following functionality has been developed in the prototype

  • Limited Monitoring and Control functionality with the hierarchy of nodes
  • LMC simulator for Dish
  • Command and response mechanism
  • Group command execution
  • Alarm detection and handling
  • Logging

The prototype incorporates the following architectural aspects

  • Model the Hierarchy of control nodes – Central node, Subarray node, Leaf node
  • Interface between TMC and Element LMC
  • Implementation of Alarm Handler using Elettra
  • Use of base classes for development of control nodes and LMC simulators

The following diagram represents the architecture of the prototype

Development of SKA LMC Base Classes:

The development of the SKA LMC Base classes work, jointly taken up by the South African and Indian Team, before TM CDR was halted ahead of the TM CDR Review meeting. The Indian Agile Team resumed the work as part of the Bridging activity. Following has been the scope and objective of the activity

  • Review and migration of base class repository to the SKA repository
  • Setting up Docker, Ansible and Pytest environment
  • Development of SKALogger and SKAAlarmHandler base classes
    • SKALogger
      • Development of SKALogger class
      • Send logs to different targets: Element Logger, Central Logger and Syslog
      • Exploration of setting different logging levels for different targets
      • Set logs at different levels
      • Exploration of Python logging and Tango Logging services for possible merging
    • SKAAlarmHandler
      • Exploration of Elettra Alarm Handler
      • Analysis of SKAAlarmHandler requirements, mapping to Elettra features
      • Configuration of alarms / alarm rule management using Elettra and PANIC

TANGO HDB++ Archiver Testing:

The primary objective of this activity was to understand the scalability of TANGO Archiving Solution (HDB++ along with Cassandra). It is beneficial to understand the behavior of HDB++ archiver as various parameters related to archival (number of monitoring points, data types of monitoring points, the frequency of archive event) are modified. The following activities were undertaken

  • Characterization of Tango Archiver instance (number of monitoring points and frequency of archiver events it can handle efficiently)
  • Stability Testing
  • Burst Load Testing
  • Determine horizontal scalability of HDB++ event archivers. e. scaling up by deploying multiple instances of event archivers
  • Scalability of Cassandra Database Cluster

SKA RTEs Visit to India:

Marco Bartolini, the Release Train Engineer (RTE), visited NCRA, India in the month of November for four days. During his visit, he had extended discussions with the Indian Agile team on the way of working as part of the SAFe and SKA platforms to be used for software development. He visited Indian industry partners (TCS and PSL) and had useful discussions with them. The Indian industry partners made presentations highlighting their capabilities in the areas of SAFe implementation and other areas of software development during construction. He also visited GMRT Telescope on one of the days.

SAFe Workshop:

One of the Indian industry partners organised two-day SAFe workshop for the Indian Agile Team ahead of the first PI Planning meeting. The aim of the workshop was to give an overview of the SAFe methodology and was mainly focused on getting the team ready for the PI meeting. The mock PI planning session conducted by the consultant found to be very useful.


Report provided by the TM consortium