New era of data analysing. Product Presentation 2014

Size: px
Start display at page:

Download "New era of data analysing. Product Presentation 2014"

Transcription

1 New era of data analysing Product Presentation 2014

2 Trace Analyzer overview Trace Analyzer is a tool for post analyzing of massive data amounts from various different sources. It reduces the time used for the analyzing with powerful internal functions. Trace Analyzer has the ability to read and organize the log data based on time stamps and log types into a single view saving user the time consuming effort of doing the organizing manually. Time based alignment quickly shows all of the system components events at an exact point in time giving the user a full view on the system. This tool enables a quicker transition into error correction and implementation phases.

3 Application areas Trace Analyzer is a tool for post analyzing of massive data amounts from various different sources and improving productivity in systems development. Typical characteristics of the customer are - Product companies more typically than service companies - Company has strong software R&D - Software code base is critical for product functionality - Software under testing provides traces as log files - Company is utilizing development tools that are not owned by its customer or project customer.

4 Typical users or user groups In software or systems engineering organization, the users can be Software testing finding relevant information, sorting out unnecessary Software developers - make software error correction easier System experts end-to-end system level view for functionality and issues Radio network specialists data mining of the system Test automation lowering R&D costs by automated testing Software designers improved understanding for correcting bug Test managers - overview of test results

5 Phases of data analyzing As working with data from various sources, most of the time is usually consumed in pre-analyzing and analyzing phase. Lot of time is spend on looking same timespan on all of the files. One of the most time consuming phase is information transfer/sharing in-/out-side of the team. Reduce average data analyzing time up to 40% Pre-analyzing Analyzing Corrective actions Verification Deployment Trace Analyzer can provide meaningful savings in data analyzing time.

6 Normal pain points in R&D Major issues on problem solving in software engineering 1. synchronize multiple log-files with each others 2. removing (must do) routine work 3. sharing knowledge and information in-/out-side of the team 4. exclude earlier known problems automatically Sharing of knowledge / documentation of findings Automation of data analysis Data Collection Pre-analyzing Analyzing Corrective actions Verification Deployment Data files synchronization

7 Example of documenting findings

8 Examble of summing up Analyze summary can be exported to customer specific format.

9 Example of graphs & Plots Visualize raw trace data to get the big picture of system

10 Product variants Trace Analyzer - Pro is the most versatile version of the Trace Analyzer providing all of the tools that are needed for effectively analyzing. Demographic: Professionals with the responsibility of heavy, detailed analyzing and problem solving responsibilities. Trace Analyzer - Lite is a affordable version of the Trace Analyzer. Demographic: Professionals with the responsibility of light analyzing and pre-analyzing the log files. Trace Analyzer CLI is a fully automated version of Trace Analyzer. Demographic: Professionals responsible of automated testing and automated analyzing.

11 Use Case of vendor 1 Easy and complete view to system issues Without Trace Analyzer : 1. convert and filter files to human readable format using a script or some other tool 2. open all log files with a text editor or some proprietary tool 3. analyze each log source separately, sometimes even by different people 4. possible time offsets between logs have to be manually recorded, difficult to synchronize -> manual pre-processing -> manual synchronization -> impossible to see the whole picture With Trace Analyzer: 1. automatically process log files from different sources and combine them into one view and time reference 2. events between system elements can be synchronized and all sources investigated together with a common time reference 3. e.g. multicore log can be automatically filtered and each core analyzed separately Less manual tasks Overview of the system

12 Use Case of vendor 1 Trace Analyzer supports Wireshark log formats Without Trace Analyzer : 1. Wireshark used as primary BTS side log analysis tool 2. Aeroflex TM500 logs analyzed manually Manual and difficult analysis of log files in different formats and different tools With Trace Analyzer: 1. same wireshark logs can be opened + additionally filtered into multiple views which Wireshark cannot do 2. Aeroflex TM500 support available gives user the ability to combine all Ue's into one, continuous log file for more quality analysis

13 Use Case of vendor 2 Continuous Integration (CI): Find unwanted events from a set of log files generated by a regression test run Without Trace Analyzer : 1. use grep or other similar tool to scan log files for an unwanted event, counting the lines 2. show which line in which file has an unwanted event 3. setup CI to automatically trigger a failed test case if an unwanted event is encountered 4. further analysis difficult -> failed CI cases study started from scratch With Trace Analyzer: 1. run a pre-defined analysis after each continuous integration test (CI will run automatically) 2. CLI automatically check for lines containing an unwanted event and reports the total count 3. detailed root cause analysis can be done with Trace Analyzer which has all the an unwanted events highlighted automatically Better information Information sharing

14 Use Case of vendor 2 Performance analysis: Record hardware usage and automatically indicate values exceeding certain threshold Without Trace Analyzer : 1. grep and filter values from original files to a temporary file 2. convert file format 3. import data into a spreadsheet application or other data analysis tool 4. plot or filter data to locate unwanted behavior 5. if an unwanted value is found, back trace into the original file to locate the time when this occurred With Trace Analyzer: 1. run pre-define state machine which intelligently collects resource usage for each core separately 2. Trace Analyzer automatically generate comment at each point where pre-defined threshold value is exceeded 3. possibility to do long term statistical follow-up of key performance indicators Quality control and performance control -> quality / performance information hard to get (usually not available)

15 Use Case of vendor 3 Track misbehaving Ue: Investigate timing and parameters of multiple simultaneous calls Without Trace Analyzer : 1. look at the combined log file 2. every Ue has to be separately manually analyzed 3. track of each of the Ue s has to be done manually -> impossible to see the whole picture -> tracking misbehaving Ue s difficult With Trace Analyzer: 1. automatically separate each Ue to an own view 2. compare timing of calls and other performance, quickly pinpoint misbehaving Ues 3. easily see behavior of all Ue s simultaneously See the whole picture / removing waste

16 Use Case of vendor 3 Effective visualizing of system issues Without Trace Analyzer : 1. convert and filter files to human readable format using a script or some other tool 2. using script or by manual; collect needed information from several different file. 3. import to spreadsheet software 4. using e.g. macro; calculate incidence of events on selected windowing 5. plot incidence into graph 6. if glitch found, manually trace back to raw data to locate the source of glitch With Trace Analyzer: 1. create find, which will hit on desired keyword. 2. open summary and check timeline graph so see incidence of find 3. if glitch found, synchronization of raw data can be done from timeline graph directly New information available, more reliable decisions making -> Extreme arduous to do. -> usually not done, unless it s necessary (problems arise)

17 Trace Analyzer roadmap 04/2014 Major upgrade releases (feature sets) once a year. Minor updates and fixes in the monthly releases. 3.x TA x Trace Analyzer 2.x commercial release Maintenance 1.x Trace Analyzer Pro & CLI first commercial release Maintenance 4/13 7/13 10/13 1/14 4/14 7/14 10/14 1/15

18 Product releases Release 1.0 1st commercial release for customers Release 2.0 File formats configurable with JSON Stability and usability improvements New, simplified welcome dialog Release 2.1 Support for common log formats Stability and usability improvements Global scrollbar Release 2.2 Support for hierarchical logs Preprosessing criteria

19 Features under planning 2.3 release (05/2014) hierarchical variables variable data table view preprocessing scripts, e.g. open and parse binary TTI trace directly delayed start and defined maximum duration for state machines 2.4 release (06/2014) test report based on format (under spefication) cumulative counters for events average event count over time period specification upper and lower limits for gathered values

20 Summary of benefits Find the bugs or issues easier and quicker Sort out unnecessary data to find the relevant information Improve understanding or visibility to system under testing Benefit your co-worker s work with Trace Analyzer Reduce or removes manual tasks or routines Enable efficient communication between teams or individuals even across organization boundaries Automate the manual tasks into repeatable sequences Enable efficient regression testing Shorten the development time Improved progress Higher quality of decision making Less s, less hassle

21 Further information For more information: Ville-Veikko Keränen Product Manager +358 (0) Juhani Leppänen Director Sales and Business development +358 (0)