Hi Marco,
several questions:
- how do you open the file, via TFile::Open("rfio:/shift/na49i06/....") ?
- if you start with this file you can read it to the end ?
It might be a temporary network problem which RFIO does not gracefully handle
which in return causes the ReadBuffer error in TFile (the TFile::ReadBuffer
error means that a low level read() has failed).
Could you try with the latest ROOT which has been linked with the latest RFIO
library.
Cheers, Fons.
Marco van Leeuwen wrote:
>
> Hi All,
>
> I repeatedly get a reading error while reading root-files. I don't know
> the exact cause of this, but it only occurred after our system was
> 'upgraded' from RH 5.1 to RH 6.2. First I thought that it might be
> connected to my own code, so I dressed down the program to a minimum, but
> it still occurs. I read (NA49) mini-DST from a disk which is accessible
> through rfio. Below is a snippet of the output, including the error
> message:
>
> Event 1413 3085(110000) (collected Parts: 125506638)
> Real time 5.403e+04 CPU 9.111e+03
> SysError in <TFile::ReadBuffer>: error reading from file
> /shift/na49i06/data1/root49-mdst/00B/t49run1413.3.root (Input/output
> error)
> Error in <TBranch::GetBasket>: File problem at address:36217952, basket
> seekkey=0, branch:fPrimaryParticles.fZFirst
> SysError in <TFile::ReadBuffer>: error reading from file
> /shift/na49i06/data1/root49-mdst/00B/t49run1413.3.root (Input/output
> error)
> Error in <TBranch::GetBasket>: File problem at address:36229864, basket
> seekkey=0, branch:fPrimaryParticles.fXLast
> Event 1413 3586(110500) (collected Parts: 126060441)
>
> As you can see the programme has been reading for 54030 s, i.e. 15 hours,
> so I don't expect any problems with the afs tokens. (Note that the 15
> hours reading is not a single file, but a list of files.) Although another
> 'local' error is not excluded. Could anyone please give me an indication
> of what these error messages mean??
>
> I'm still using ROOT 2.23/12, but will try the same stuff with the newest
> ROOT version, although I'm not aware of any changes which would affect
> the observed behaviour. Please be aware that a new 'test' each time takes
> many hours, because the error generally occurs after a long period of
> reading....
>
> Thanks in advance,
>
> Marco van Leeuwen.
--
Org: CERN, European Laboratory for Particle Physics.
Mail: 1211 Geneve 23, Switzerland
E-Mail: Fons.Rademakers@cern.ch Phone: +41 22 7679248
WWW: http://root.cern.ch/~rdm/ Fax: +41 22 7677910
This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:31 MET