KAGRA/Subgroups/CAL/GstLAL/tutorials/step11

Step 11: Write channel data to shared memory

For the initialization and running the pipeline, please refer step1
Access frame cache file by specifying frame_cache and instrument as in step2
Define a frame multiplexer(mux) to write channels into a frame as in step3
The only difference from (step3)writing to a frame file is use gds_lvshmsink instead of mkframecppfilesink

   1 pp.mkgeneric(pipeline, mux, "gds_lvshmsink", sync=False, async=False,\
   2              shm_name = options.shared_memory_partition, num_buffers=10,\
   3              blocksize = options.frame_size*frame_duration*frames_per_file,\
   4              buffer_mode = options.buffer_mode)

The python scripts are available at git repository

> kagra-cal/Observation/Phase1/tutorials/shmo --help
Usage: shmo [options]

Options:
  -h, --help            show this help message and exit
  --cache=cache         Set the name of the LAL cache listing the frame files
  --gps-start-time=seconds
                        Set the start time of the segment to analyze in GPS
                        seconds.
  --gps-end-time=seconds
                        Set the end time of the segment to analyze in GPS
                        seconds.
  --ifo=name            Name of the IFO (default= K1)
  --channel=name        Set channel name(s) to extract (default= CAL-
                        CS_PROC_C00_STRAIN_DQ,CAL-CS_PROC_MICH_DELTA_TM_DQ
                        ,CAL-CS_PROC_MICH_RESIDUAL_DQ,CAL-
                        CS_PROC_ERR_STRAIN_DQ,CAL-CS_PROC_CTRL_STRAIN_DQ,LSC-
                        MICH1_IN1_DQ,LSC-MICH1_OUT_DQ)
  --format=format       Set input channel(s) format (default= F64LE)
  --ch-rate=rate        Set input channel data rate (default= 16384)
  --frame-duration=seconds
                        Set the number of seconds for each frame. (Default =
                        4)
  --frame-size=bytes    Approximate size in bytes of frame file images; used
                        when writing to shared memory.  (Default=500000)
  --shared-memory-partition=name
                        Set the name of the shared memory partition to write
                        to (default = K1).
  --buffer-mode=number  Set the buffer mode for the lvshmsink element.
                        (Default=2)

Before running the script, make sure that shared memory with the same name is not running, and if so kill it with smkill

> smlist
  id      name        nBuf   lBuf     nTotal
  32  K1                1     70000        32
> smkill K1
> smlist
  id      name        nBuf   lBuf     nTotal

This is an example to extract channel data from a cache file, k1.cache and write to a shared memory

> kagra-cal/Observation/Phase1/tutorials/shmo --cache=k1.cache
GPS:  1209288320 1209288352
[('K1', 'CAL-CS_PROC_C00_STRAIN_DQ'), ('K1', 'CAL-CS_PROC_MICH_DELTA_TM_DQ'), ('K1', 'CAL-CS_PROC_MICH_RESIDUAL_DQ'), ('K1', 'CAL-CS_PROC_ERR_STRAIN_DQ'), ('K1', 'CAL-CS_PROC_CTRL_STRAIN_DQ'), ('K1', 'LSC-MICH1_IN1_DQ'), ('K1', 'LSC-MICH1_OUT_DQ')]
seeking GPS start and stop times ...
setting pipeline state to playing ...
set to playing successfully
running pipeline ...
progress_src_CH6 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_src_CH7 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_src_CH1 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_src_CH5 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_src_CH4 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_src_CH2 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_src_CH3 (00:00:01): 1209288352 / 32 seconds (3779026100.0 %)
progress_sink_K1 (00:00:06): 1209288324 seconds
progress_sink_K1 (00:00:12): 1209288336 seconds
progress_sink_K1 (00:00:17): 1209288344 seconds
progress_sink_K1 (00:00:19): 1209288352 / 32 seconds (3779026100.0 %)

Check if the shared memory partition is running

> smlist
  id      name        nBuf   lBuf     nTotal
  32  K1                10   2000000        18
smdump K1

            Event Manager Partition - K1

Global Control Data
-------------------
  version:           1.03
  status:            scavage
  global use count:  3
  global sem id:     4489216
  # buffers:         10
  buffer length:     2000000
  first full buffer: 0
  last full buffer:  6
  first free buffer: 7
  #consumers:        1
  consumer sem id:   4521985   4554754   4587523   4620292
  consumer mask:     1

Buffer Status
-------------
 id status reserved seen  use   ldata   link  t-mask   count
  0    1        0      1    0  1663469    1       -1       1
  1    1        0      1    0  1638044    2       -1       1
  2    1        0      1    0  1635658    3       -1       1
  3    1        0      1    0  1677513    4       -1       1
  4    1        0      1    0  1660442    5       -1       1
  5    1        0      1    0  1671079    6       -1       1
  6    1        0      1    0  1626121   -1       -1       1
  7    1        0      0    0        0    8        0       0
  8    1        0      0    0        0    9        0       0
  9    1        0      0    0        0   -1        0       0

Consumer Status
---------------
 id  pid  max-bf   trig   #segs   read  flags
  0 14969    0  ffffffff     0        6 wait

At the same time if you run another GstLAL script, shmi as in step 10 you will receive the C00 strain channel data

> kagra-cal/Observation/Phase1/tutorials/shmi
setting pipeline state to playing ...
set to playing successfully
1209288320.000000000    3.96861363216594e-11
1209288321.000000000    9.805171258259193e-11
1209288322.000000000    -6.216620274113183e-11
1209288323.000000000    -3.847490728792202e-11
1209288324.000000000    -1.028217927467878e-10
1209288325.000000000    7.444918292964786e-11
1209288326.000000000    1.882813627229751e-10
1209288327.000000000    -2.34722963288192e-10
progress_src_input (00:00:06): 1209288328 seconds
1209288328.000000000    3.754893097840384e-12
1209288329.000000000    -1.839911834000674e-10

KAGRA/Subgroups/CAL/GstLAL/tutorials/step11 (last edited 2018-08-21 17:39:32 by SadakazuHaino)