Author: Bourtembourg, R.
Paper Title Page
WED3O04 HDB++: A New Archiving System for TANGO 652
 
  • L. Pivetta, C. Scafuri, G. Scalamera, G. Strangolino, L. Zambon
    Elettra-Sincrotrone Trieste S.C.p.A., Basovizza, Italy
  • R. Bourtembourg, J.L. Pons, P.V. Verdier
    ESRF, Grenoble, France
 
  The TANGO re­lease 8 led to sev­eral en­hance­ments, in­clud­ing the adop­tion of the Ze­roMQ li­brary for faster and light­weight event-dri­ven com­mu­ni­ca­tion. Ex­ploit­ing these im­proved ca­pa­bil­i­ties, a high per­for­mance, event-dri­ven archiv­ing sys­tem writ­ten in C++ has been de­vel­oped. It in­her­its the data­base struc­ture from the ex­ist­ing TANGO His­tor­i­cal Data Base (HDB) and in­tro­duces new stor­age ar­chi­tec­ture pos­si­bil­i­ties, bet­ter in­ter­nal di­ag­nos­tic ca­pa­bil­i­ties and an op­ti­mized API. Its de­sign al­lows stor­ing data into tra­di­tional data­base man­age­ment sys­tems such as MySQL or into NoSQL data­base such as Apache Cas­san­dra. This paper de­scribes the soft­ware de­sign of the new HDB++ archiv­ing sys­tem, the cur­rent state of the im­ple­men­ta­tion and gives some per­for­mance fig­ures and use cases.  
slides icon Slides WED3O04 [1.392 MB]  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
WEM310 How Cassandra Improves Performances and Availability of HDB++ Tango Archiving System 685
 
  • R. Bourtembourg, J.L. Pons, P.V. Verdier
    ESRF, Grenoble, France
 
  The TANGO re­lease 8 led to sev­eral en­hance­ments, in­clud­ing the adop­tion of the Ze­roMQ li­brary for faster and light­weight event-dri­ven com­mu­ni­ca­tion. Ex­ploit­ing these im­proved ca­pa­bil­i­ties, a high per­for­mance, event-dri­ven archiv­ing sys­tem, named Tango HDB++*, has been de­vel­oped. Its de­sign gives the pos­si­bil­ity to store archiv­ing data into Apache Cas­san­dra: a high per­for­mance scal­able NoSQL dis­trib­uted data­base, pro­vid­ing High Avail­abil­ity ser­vice and repli­ca­tion, with no sin­gle point of fail­ure. HDB++ with Cas­san­dra will open up new per­spec­tives for TANGO in the era of big data and will be the start­ing point of new big data an­a­lyt­ics/data min­ing ap­pli­ca­tions, break­ing the lim­its of the archiv­ing sys­tems which are based on tra­di­tional re­la­tional data­bases. This paper de­scribes the cur­rent state of the im­ple­men­ta­tion and our ex­pe­ri­ence with Apache Cas­san­dra in the scope of the Tango HDB++ pro­ject. It also gives some per­for­mance fig­ures and use cases where using Cas­san­dra with Tango HDB++ is a good fit.
* HDB++ project is the result of a collaboration between the Elettra synchrotron (Trieste) and the European Radiation Synchrotron Facility (Grenoble)
 
slides icon Slides WEM310 [1.897 MB]  
poster icon Poster WEM310 [2.446 MB]  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)