ALICE O2 Project
Search

Detector commissioning – ITS

This page is the entry point for the documentation of the ITS readout commissioning setup.

Hardware setup

The readout test setup consists of the following equipment: (as of 08/2019)

  • 10 FLP servers (on loan from HLT): flpits0-10
  • 16 CRUs (owned by ITS), 2 installed in each machine.
  • Two disk arrays (on loan from DAQ) providing a disk pool, connected to each FLP by FiberChannel, with one local 8TB volume mounted on each FLP in /data. This setup allows recording data at ~1GB/s on each FLP.
  • Two general purpose servers (on loan from DAQ) to run the shared services. At the moment, only flpits-srv-0 is used (top of the rack), providing in particular the NFS shared file system accessible from all FLPs in /shareFS (800+ GB).
  • Network switch (on loan from DAQ), 1Gb network (between machines, and also uplink on CERN GPN).
  • Power Distribution Unit (on loan from DAQ).

Equipment is located in a rack in 23-R-020, close to the ITS clean room in building 167.

There is one additionnal FLP server (on loan from HLT) flpits11 with 2 CRUs (owned by ITS) located in 301-01-029.

Using the setup

The example commands below login to the machine, setup the runtime environment, list the available CRU cards, and readout 40GB of data from link 0 on CRU end-point connected to PCI slot 02:00.0, using a buffer of 2GB and a superpage size of 2MB.

  • ssh its@flpits1
  • aliswmod enter Readout/v1.3.6-1
  • roc-list-cards
  • roc-bench-dma –id=02:0.0 –buffer-size=2Gi –superpage-size=2Mi –links=”0″ –bytes=40Gi –verbose –no-errorche

ReadoutCard utlitities documentation: https://github.com/AliceO2Group/ReadoutCard/

Readout.exe documentation: https://github.com/AliceO2Group/Readout/blob/master/doc/README.md

A dedicated instance of the ALICE logbook has been setup: https://alice-logbook.cern.ch/its-run3

(currently base on Run2 logbook, to be moved to O2 logbook when available)

Knowledge base

  • There are two user accounts:
    • an its account with standard privileges.
    • an its-admin account with sudo privileges to become root, in case admin operations are needed.
  • In case of rack power cut:
    • the FLPs may need to be rebooted manually (they should be started only after the disk array has booted, or they get lock trying to boot from the FC card – this is a known limitation without solution).
    • the NFS service may need to be restarted manually (the access control is based on IP, but the FLPs have dynamic IPs: they get an access denied from the server if the hostname – IP mapping has changed after NFS server started):
      1. as root, on flpits-srv-0: service nfs restart
      2. as root, on flpits-[0-11] (if /shareFS is not mounted): mount /shareFS
  • The management interface of the disk arrays is accessible by launching SMclient as root on flpits-srv-0

Support

Technical documentation on various O2 components is available on this website (https://alice-o2-project.web.cern.ch/). Ad-hoc links have been included above, where necessary.

There is a discussion channel to enquire about O2 software and hardware usage: https://alice-talk.web.cern.ch/

To report bug and issues, please use JIRA (and specify ITS in the components field).

ITS-O2 contact: sylvain.chapeland@cern.ch / 75001

Loading…