The PAWC common block is dimensiooned to 2000,000 words (8 Mbytes).
This should be largely sufficient for the conversion unless:
- by mistake, you create an Hbook memory resident ntuple > 8 Mbytes
- You have a ntuple with a huge number of columns and have specified
a very large buffer size (unlikely).
Check hypothesis 1 first.
You can increase the size of the PAW common block by updating the statement
#define PAWC_SIZE 2000000
at the beginning of h2root.cxx and remake h2root.
Rene Brun
Mei Wen wrote:
>
> Part 1.1 Type: Plain Text (text/plain)
> Encoding: base64
Dear rooters,
What's the limit of the max .hbook file size when we use h2root? (or limit
of statistics?) Now in KLOE we are taking long runs data about (100nb-1) for
each run. It happens that for some runs when the .hbook file >10MB the h2root is
not able to convert the file to .root successfully. Some strange message like
following:
(HI1E. 2 1882077( ) SY/US/IO 0040/00000/0003 NL/NS/ND 0/
0/ 60 N/U/O/@O 0/ 1882232/ 1882150/ 1882077
(HCO1. 1 1882150( ) SY/US/IO 0041/00000/31A3 NL/NS/ND 2/
2/ 70 N/U/O/@O 1882077/ 1882232/ 1882231/ 1882150
(HID1. 7000 1882232( ) SY/US/IO 0041/00010/2D15 NL/NS/ND 1/
1/ 21 N/U/O/@O 0/ 1995059/ 1995057/ 0
** LINKS 1882150(HCO1
At the end the file .root might not be well closed. Because when I opened the
.root file afterwords I got warning err message
TFile:TFile> file probably not closed. trying to recover.
TFile:Recover> successfully recovered 3 keys.
Could you give me an explanation? I can see the histograms after the file
opened. But I am not sure the file has been converted correctly. Thanks a lot.
By the way, thanks to Brun. After I redefined Directory instead of TFile as you
wrote, now every pieces works properly!:)
Cheers
Meiwen
This archive was generated by hypermail 2b29 : Tue Jan 01 2002 - 17:51:05 MET