ICALEPCS2017 - List of Authors (Kurepin, A.N.)

Paper	Title	Page
MODPL07	How Low-Cost Devices Can Help on the Way to ALICE Upgrade	114
	O. Pinazza INFN-Bologna, Bologna, Italy A. Augustinus, P.M. Bond, P.Ch. Chochula, A.N. Kurepin, M. Lechman, J.L. LÃ¥ng, O. Pinazza CERN, Geneva, Switzerland A.N. Kurepin RAS/INR, Moscow, Russia
	The ambitious upgrade plan of the ALICE experiment expects a complete redesign of its data flow after the LHC shutdown scheduled for 2019, for which new electronics modules are being developed in the collaborating institutes. Access to prototypes is at present very limited and full scale prototypes are expected only close to the installation date. To overcome the lack of realistic HW, the ALICE DCS team built small-scale prototypes based on low-cost commercial components (Arduino, Raspberry PI), equipped with environmental sensors, and installed in the experiment areas around and inside the ALICE detector. Communication and control software was developed, based on the architecture proposed for the future detectors, including CERN JCOP FW and ETM WINCC OA. Data provided by the prototypes has been recorded for several months, in presence of beam and magnetic field. The challenge of the harsh environment revealed some insurmountable weaknesses, thus excluding this class of devices from usage in a production setup. They did prove, however, to be robust enough for test purposes, and are still a realistic test-bed for developers while the production of final electronics is continuing.
	Talk as video stream: https://youtu.be/utSHzqk44hQ
	Slides MODPL07 [9.016 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICALEPCS2017-MODPL07
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUMPL09	Challenges of the ALICE Detector Control System for the LHC RUN3	323
	P.Ch. Chochula, A. Augustinus, P.M. Bond, A.N. Kurepin, M. Lechman, J.L. LÃ¥ng, O. Pinazza CERN, Geneva, Switzerland A.N. Kurepin RAS/INR, Moscow, Russia M. Lechman IP SAS, Bratislava, Slovak Republic O. Pinazza INFN-Bologna, Bologna, Italy
	The ALICE Detector Control System (DCS) provides its services to the experiment for 10 years. It ensures uninterrupted operation of the experiment and guarantees stable conditions for the data taking. The decision to extend the lifetime of the experiment requires the redesign of the DCS data flow. The interaction rates of the LHC in ALICE during the RUN3 period will increase by a factor of 100. The detector readout will be upgraded and it will provide 3.4TBytes/s of data, carried by 10 000 optical links to a first level processing farm consisting of 1 500 computer nodes and ~100 000 CPU cores. A compressed volume of 20GByte/s will be transferred to the computing GRID facilities. The detector conditions, consisting of about 100 000 parameters, acquired by the DCS need to be merged with the primary data stream and transmitted to the first level farm every 50ms. This requirement results in an increase of the DCS data publishing rate by a factor of 5000. The new system does not allow for any DCS downtime during the data taking, nor for data retrofitting. Redundancy, proactive monitoring, and improved quality checking must therefore complement the data flow redesign.
	Slides TUMPL09 [1.773 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICALEPCS2017-TUMPL09
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPHA042	ADAPOS: An Architecture for Publishing ALICE DCS Conditions Data	482
	J.L. LÃ¥ng, A. Augustinus, P.M. Bond, P.Ch. Chochula, A.N. Kurepin, M. Lechman, O. Pinazza CERN, Geneva, Switzerland A.N. Kurepin RAS/INR, Moscow, Russia M. Lechman IP SAS, Bratislava, Slovak Republic O. Pinazza INFN-Bologna, Bologna, Italy
	ALICE Data Point Service (ADAPOS) is a software architecture being developed for the Run 3 period of LHC, as a part of the effort to transmit conditions data from ALICE Detector Control System (DCS) to GRID, for distributed processing. ADAPOS uses Distributed Information Management (DIM), 0MQ, and ALICE Data Point Processing Framework (ADAPRO). DIM and 0MQ are multi-purpose application-level network protocols. DIM and ADAPRO are being developed and maintained at CERN. ADAPRO is a multi-threaded application framework, supporting remote control, and also real-time features, such as thread affinities, records aligned with cache line boundaries, and memory locking. ADAPOS and ADAPRO are written in C++14 using OSS tools, Pthreads, and Linux API. The key processes of ADAPOS, Engine and Terminal, run on separate machines, facing different networks. Devices connected to DCS publish their state as DIM services. Engine gets updates to the services, and converts them into a binary stream. Terminal receives it over 0MQ, and maintains an image of the DCS state. It sends copies of the image, at regular intervals, over another 0MQ connection, to a readout process of ALICE Data Acquisition.
	Poster TUPHA042 [0.686 MB]
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-ICALEPCS2017-TUPHA042
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

How Low-Cost Devices Can Help on the Way to ALICE Upgrade

114

O. Pinazza
INFN-Bologna, Bologna, Italy
A. Augustinus, P.M. Bond, P.Ch. Chochula, A.N. Kurepin, M. Lechman, J.L. LÃ¥ng, O. Pinazza
CERN, Geneva, Switzerland
A.N. Kurepin
RAS/INR, Moscow, Russia

The ambitious upgrade plan of the ALICE experiment expects a complete redesign of its data flow after the LHC shutdown scheduled for 2019, for which new electronics modules are being developed in the collaborating institutes. Access to prototypes is at present very limited and full scale prototypes are expected only close to the installation date. To overcome the lack of realistic HW, the ALICE DCS team built small-scale prototypes based on low-cost commercial components (Arduino, Raspberry PI), equipped with environmental sensors, and installed in the experiment areas around and inside the ALICE detector. Communication and control software was developed, based on the architecture proposed for the future detectors, including CERN JCOP FW and ETM WINCC OA. Data provided by the prototypes has been recorded for several months, in presence of beam and magnetic field. The challenge of the harsh environment revealed some insurmountable weaknesses, thus excluding this class of devices from usage in a production setup. They did prove, however, to be robust enough for test purposes, and are still a realistic test-bed for developers while the production of final electronics is continuing.

Talk as video stream: https://youtu.be/utSHzqk44hQ

Slides MODPL07 [9.016 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICALEPCS2017-MODPL07

Export •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUMPL09

Challenges of the ALICE Detector Control System for the LHC RUN3

323

P.Ch. Chochula, A. Augustinus, P.M. Bond, A.N. Kurepin, M. Lechman, J.L. LÃ¥ng, O. Pinazza
CERN, Geneva, Switzerland
A.N. Kurepin
RAS/INR, Moscow, Russia
M. Lechman
IP SAS, Bratislava, Slovak Republic
O. Pinazza
INFN-Bologna, Bologna, Italy

The ALICE Detector Control System (DCS) provides its services to the experiment for 10 years. It ensures uninterrupted operation of the experiment and guarantees stable conditions for the data taking. The decision to extend the lifetime of the experiment requires the redesign of the DCS data flow. The interaction rates of the LHC in ALICE during the RUN3 period will increase by a factor of 100. The detector readout will be upgraded and it will provide 3.4TBytes/s of data, carried by 10 000 optical links to a first level processing farm consisting of 1 500 computer nodes and ~100 000 CPU cores. A compressed volume of 20GByte/s will be transferred to the computing GRID facilities. The detector conditions, consisting of about 100 000 parameters, acquired by the DCS need to be merged with the primary data stream and transmitted to the first level farm every 50ms. This requirement results in an increase of the DCS data publishing rate by a factor of 5000. The new system does not allow for any DCS downtime during the data taking, nor for data retrofitting. Redundancy, proactive monitoring, and improved quality checking must therefore complement the data flow redesign.

Slides TUMPL09 [1.773 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICALEPCS2017-TUMPL09

Export •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPHA042

ADAPOS: An Architecture for Publishing ALICE DCS Conditions Data

482

J.L. LÃ¥ng, A. Augustinus, P.M. Bond, P.Ch. Chochula, A.N. Kurepin, M. Lechman, O. Pinazza
CERN, Geneva, Switzerland
A.N. Kurepin
RAS/INR, Moscow, Russia
M. Lechman
IP SAS, Bratislava, Slovak Republic
O. Pinazza
INFN-Bologna, Bologna, Italy

ALICE Data Point Service (ADAPOS) is a software architecture being developed for the Run 3 period of LHC, as a part of the effort to transmit conditions data from ALICE Detector Control System (DCS) to GRID, for distributed processing. ADAPOS uses Distributed Information Management (DIM), 0MQ, and ALICE Data Point Processing Framework (ADAPRO). DIM and 0MQ are multi-purpose application-level network protocols. DIM and ADAPRO are being developed and maintained at CERN. ADAPRO is a multi-threaded application framework, supporting remote control, and also real-time features, such as thread affinities, records aligned with cache line boundaries, and memory locking. ADAPOS and ADAPRO are written in C++14 using OSS tools, Pthreads, and Linux API. The key processes of ADAPOS, Engine and Terminal, run on separate machines, facing different networks. Devices connected to DCS publish their state as DIM services. Engine gets updates to the services, and converts them into a binary stream. Terminal receives it over 0MQ, and maintains an image of the DCS state. It sends copies of the image, at regular intervals, over another 0MQ connection, to a readout process of ALICE Data Acquisition.

Poster TUPHA042 [0.686 MB]

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-ICALEPCS2017-TUPHA042

Export •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)