Test Equipment Utilization Measurements & Usage Optimization – Progress Report II (One Month CheckPoint)

View Github Repository Open presentation in a new window

Zz-Wang

See all presentation from Zz-Wang

Test Equipment Utilization Measurements & Usage Optimization – Progress Report II (One Month CheckPoint)

0 0

second_presentation

wzz

On Github Zz-Wang / second_presentation

Test Equipment Utilization Measurements & Usage Optimization

Progress Report II (One Month CheckPoint)

Zhizheng Wang

Novice at BUCI DUNC PC NDO GW LSV&CI ST Frame Work

Supervised by Stefan Johannesson

2015-02-20

Outline

Research and investigate existing utilization and usage tools outside and inside of Ericsson for similar products as EPG. Suggest different improvements for KPIs to measure the equipment utilizations Propose different alternatives for optimizing equipment usage Prototype technical solutions (Visualizing) for equipment utilization and usage optimization.

Schedule and Plan

Date Contents 26th Jan. Thesis Beginning 6th Feb. Monitering I 20th Feb. Monitering II 6th Mar. Monitering III 20th Mar. Midterm Report 3rd April Monitering IV 17th April Monitering V 1st May Monitering VI 15th May Monitering VII 29th May Monitering III 3rd Jun. Thesis Defense(Univ) 12 Jun. Monitering XI 22nd Jun. Thesis Closing(Ericsson)

Echo: From first progress report

Some focus are "misleading" (Portal) Redefine the project goal Overlook ECUT

Portal Progress and Expectation

  • Portal (DB) gives more broad/rough view of how the nodes are used. Examples:

Waiting Time VS Actual Run Time

How they exited (Duration TimedOut VS Removed by User)

Strategy 1: Avoid booking peak

Distribution of Booking (Start & End Time) Daily

Strategy 2: Cut down waiting time

  • Waiting time: From Created_at to Start
  • Run time: From Start to End

Strategy 3: What does the message say? (Planned text mining)

Redefining Goals

Questions:

  • How to define a utilization? When is an equipment "properly" used? -->> Study from tester point of view (Survey Planed to launch on Next Tuesday during ST Meeting, help from Magnus)
  • How to define the time interval? When it starts? When it ends? -->> Knowledge about the testing process, which test can/shall be moved to SSR-SIM (Reading "Testing IT-An off-the-shelf SW Testing Process" in progress)
  • What equipment we are looking at? More focus on SSR

Main Target: ECUT on EPG

Tools in hand (Axsel):

  • Collector module scripts (pl)
  • checking logs (from 2013-12-05)

Some potential problems:

  • script out-dated, new features couldn't catch
  • 1hr checking cycle could miss the actual usage
  • Strange Status

ECUT on EPG (CONT.)

Proposals and next step:

  • Studying on old counters (which are the most significant in detecting the usage change status)
  • Checking (updating) the counters (according to new features/new software version)
  • Studying the working principle of EPG to get to know when the counter is triggered. Maybe some other KPIs beyond counters? Could Performance Monitoring Statistics for the GGSN & PGW/ SGW help to as KPIs?

Outside of EPG

Metrics team, Wilhelm proposed two others (John Haagen & Tommy Davidsson) (Haven't reached) Contact Similar thesis worker (Denis reached) Contact other ECUT developer (tex. Haiyan Feng) Look outside of Ericsson

Something more:

  • Intensive Information on DB API
  • Getting more familiar with NX, vi, Perl...
  • Carried out a test using SSR-SIM to get some taste of it. (Mats Palm)