From: Ian Bearden (bearden@nbi.dk)
Date: Tue Jan 21 2003 - 08:18:07 EST
I don't know who Iants is, but I will forward this for Flemming, who writes from home. I have not (yet) gotten much beyond the reduction stage with any of these runs, so I do not know if there is anything 'odd' about the sequences which get written to HPSS (though I do not see a reason there should be). Still, we must be losing something like *at least* 25% of the events we take due to this problem. Cheers, Ian -------------------------------------------------------------------- | I.G. Bearden | | Niels Bohr Institute Tb 3 email: bearden@nbi.dk | | Blegdamsvej 17 phone: (+45) 35 32 53 23 | | København Ø FAX: (+45) 31 42 10 16 | | Danmark Www: http://www.nbi.dk/~bearden | -------------------------------------------------------------------- -----Original Message----- From: Flemming Videbaek [mailto:fvidebaek@optonline.net] Sent: 21. januar 2003 14:10 To: Ian Bearden Subject: Re: HPSS database problem? Hi Iants Can you forward this to the dev-l list too, since I am writing from home. Since the io-server crashes the run-server is not successful in making an end-of-run action. This results in the at least one files not being writting at the end run - among these precesily the ones that describes the numbers of events, triggers etc. In addition most likely the last raw data sequence is NOT saved to HPSS either since the file does not get the 'sequence-state' file to a state that indicates it is ready to be send. At present there is certainly no automatic way to performe this function, and would require a real expert on the DAQ code. Flemming ------------------------------------------------------ Flemming Videbaek Physics Department Brookhaven National Laboratory tlf: 631-344-4106 fax 631-344-1334 e-mail: videbaek@bnl.gov ----- Original Message ----- From: "Ian Bearden" <bearden@nbi.dk> To: <brahms-l@bnl.gov> Sent: Tuesday, January 21, 2003 6:13 AM Subject: RE: HPSS database problem? | Hi again, | I looked at some more detailed information and found that the runs which | show 0 events, but actually have some data are runs for which the IoServer | crashed. | Anyone know how to keep this from happening? | Cheers, | Ian | | -------------------------------------------------------------------- | | I.G. Bearden | | | Niels Bohr Institute Tb 3 email: bearden@nbi.dk | | | Blegdamsvej 17 phone: (+45) 35 32 53 23 | | | København Ø FAX: (+45) 31 42 10 16 | | | Danmark Www: http://www.nbi.dk/~bearden | | -------------------------------------------------------------------- | | -----Original Message----- | From: owner-brahms-l@bnl.gov [mailto:owner-brahms-l@bnl.gov]On Behalf Of | Ian Bearden | Sent: 21. januar 2003 07:46 | To: Brahms-L@Bnl. Gov | Subject: HPSS database problem? | | | Hi All, | I have noticed that for many of the recent runs, the DB says that they have | no events, but there are sequences in HPSS. Can someone who knows about | such things take a look at this? The run summary is extremely useful. In | fact, I would not have found out about this except that I decided to just | reduce from runs 7506-7516, rather than only reduce those with a reasonable | number of events... | Cheers, | Ian | | -------------------------------------------------------------------- | | I.G. Bearden | | | Niels Bohr Institute Tb 3 email: bearden@nbi.dk | | | Blegdamsvej 17 phone: (+45) 35 32 53 23 | | | København Ø FAX: (+45) 31 42 10 16 | | | Danmark Www: http://www.nbi.dk/~bearden | | -------------------------------------------------------------------- | | | | | | |
This archive was generated by hypermail 2.1.5 : Tue Jan 21 2003 - 08:16:41 EST