Error in <TBufferFile::WriteByteCount>

1

Dear experts,
There is an error when I run Corryvreckan, and it disappears if I disable [AnalysisDUT]:

Error in TBufferFile::WriteByteCount: bytecount too large (more than 1073741822)

Here is the log file:

...
|19:16:59.236|  (STATUS) [I:Clustering4D:telescope3] Initializing "Clustering4D:telescope3"
|19:16:59.270|  (STATUS) [I:Clustering4D:telescope4] Initializing "Clustering4D:telescope4" 
|19:16:59.303|  (STATUS) [I:Clustering4D:telescope5] Initializing "Clustering4D:telescope5" 
|19:16:59.339|  (STATUS) [I:Tracking4D] Initializing "Tracking4D"                           
|19:16:59.342|  (STATUS) [I:DUTAssociation:dut] Initializing "DUTAssociation:dut"
|19:16:59.342|  (STATUS) [I:AnalysisDUT:dut] Initializing "AnalysisDUT:dut"       
|19:17:06.396|  (STATUS) ========================| Event loop |========================
|19:17:06.423|  (STATUS) Ev: 0.1k Px: 1.4k Tr: 0.0k (0.02/ev) t = 990ns
|19:17:06.508|  (STATUS) Ev: 0.2k Px: 2.7k Tr: 0.0k (0.015/ev) t = 1.99us
|19:17:06.521|  (STATUS) Ev: 0.3k Px: 4.1k Tr: 0.0k (0.0167/ev) t = 2.99us
|19:17:06.534|  (STATUS) Ev: 0.4k Px: 5.5k Tr: 0.0k (0.02/ev) t = 3.99us   
|19:17:06.546|  (STATUS) Ev: 0.5k Px: 6.8k Tr: 0.0k (0.018/ev) t = 4.99us
|19:17:06.558|  (STATUS) Ev: 0.6k Px: 8.2k Tr: 0.0k (0.0167/ev) t = 5.99us
|19:17:06.570|  (STATUS) Ev: 0.7k Px: 9.6k Tr: 0.0k (0.0143/ev) t = 6.99us 
|19:17:06.582|  (STATUS) Ev: 0.8k Px: 11.0k Tr: 0.0k (0.0125/ev) t = 7.99us
|19:17:06.594|  (STATUS) Ev: 0.9k Px: 12.4k Tr: 0.0k (0.0111/ev) t = 8.99us 
|19:17:06.606|  (STATUS) Ev: 1.0k Px: 13.8k Tr: 0.0k (0.011/ev) t = 9.99us  
|19:17:06.606|  (STATUS) ===================| Finalising modules |===================
|19:17:07.388|  (STATUS) [F:DUTAssociation:dut] In total, 11 clusters are associated to 11 tracks.
Error in <TBufferFile::WriteByteCount>: bytecount too large (more than 1073741822)
Error in <TBufferFile::WriteByteCount>: bytecount too large (more than 1073741822)
|19:17:37.454|  (STATUS) Wrote histogram output file to /home/corry/output/test_beam.root
|19:17:37.454|  (STATUS) ===============| Wall-clock timing (seconds) |================
|19:17:37.454|  (STATUS)           FileReader               --  0.13915s = 0.139149ms/evt
|19:17:37.454|  (STATUS)         Clustering4D : telescope0  --  0.00720s = 0.007196ms/evt
|19:17:37.455|  (STATUS)         Clustering4D : telescope1  --  0.00518s = 0.005175ms/evt
|19:17:37.455|  (STATUS)         Clustering4D : telescope2  --  0.00515s = 0.005146ms/evt
|19:17:37.455|  (STATUS)         Clustering4D :        dut  --  0.00468s = 0.004675ms/evt
|19:17:37.455|  (STATUS)         Clustering4D : telescope3  --  0.00489s = 0.004888ms/evt
|19:17:37.455|  (STATUS)         Clustering4D : telescope4  --  0.00460s = 0.004601ms/evt
|19:17:37.455|  (STATUS)         Clustering4D : telescope5  --  0.00501s = 0.005010ms/evt
|19:17:37.455|  (STATUS)           Tracking4D               --  0.02485s = 0.024849ms/evt
|19:17:37.455|  (STATUS)       DUTAssociation :        dut  --  0.00154s = 0.001544ms/evt
|19:17:37.455|  (STATUS)          AnalysisDUT :        dut  --  0.00147s = 0.001468ms/evt
|19:17:37.455|  (STATUS) ==============================================================

ROOT version 6.25/01
Corryvreckan version v2.0+303^gb7163ba~dirty

Do you have any idea about the cause of this error?
Thanks.

2

Dear Yang,
thanks for reporting.
I might have seen it before - could you check if it only occurs when running the module [Correlations] or [DUTAssociation] by commenting out all subsequent module?
Tracing down which module it arises from would help a lot to see where it comes from.

We should try to fix it, but an in important question is:
Does this error affect your results, i.e. does it mean you get non-sense results?
Do you notice anything weird in your final results or could this only be related to the filling on one particular histogram?

Cheers, Jens

3

Dear Jens,
I haven’t noticed weird results though.
I just tried a few times. This error only occurs when running the module [AnalysisDUT].
According to this and this on the ROOT forum, the error might arise from this line.

4

Dear Yang,

that sounds reasonable.
Could it be that you’re using a rather large value for either of the following parameters:

  • n_timebins_
  • time_binning_
    resulting in histograms with too many bins?

Or does it only happen when using

  • correlations_ = true
    which creates a bunch of additional histograms?

Cheers, Jens

1 Like
5

Dear Jens,
I just changed the n_time_bins value from the default 20000 to 200, then the error is gone.
Thanks so much!

6

Glad I could help. Maybe we should decrease the default value to avoid this error in the future.

Cheers, Jens