average record size? standard deviation from average on record size? Wally Terhune U2 Support Architect IBM Information Management 4700 South Syracuse Street, Denver, CO 80237 Tel: 303.773.7969 Fax: 303.773.5915 [EMAIL PROTECTED]
----- Forwarded by Wally Terhune/Denver/IBM on 05/11/2006 03:36 PM ----- "Baakkonen, Rodney A (Rod) 46K" To <Rodney.Baakkonen "'u2-users@listserver.u2ug.org'" @cignabehavioral. <u2-users@listserver.u2ug.org> com> cc Sent by: [EMAIL PROTECTED] Subject stserver.u2ug.org RE: [U2] Unidata logging of BASIC Runtime errors 05/11/2006 03:16 PM Please respond to [EMAIL PROTECTED] er.u2ug.org The minimum modulo is not that bad. The current mod is 529447, and the minimum 482029. Primary File Head (dat001): modulo --------------> 122069 hashtype ------------> 96 version -------------> 3 blksiz --------------> 15 (16384) idxinf --------------> 10 bproof --------------> 1 multi-over file -----> 0 rm ------------------> 255 prt_at --------------> 192 nullv --------------> 129 beg_modulo ----------> 0 file_limit ----------> 2000000000 Overflow File Head (over001): modulo --------------> 0 (not used) hashtype ------------> 0 (not used) block size ----------> 0 (1024) (not used) next_block ----------> 0 (block 0) block_count ---------> 0 max_blk_cnt ---------> 4077 file_logical_end ----> -2070724608 current_modulo ------> 529447 min_modulo ----------> 482029 split_factor --------> 95 merge_factor --------> 40 grp_fre_blk_cnt -----> 5 file_limit ----------> 2000000000 total_free_blk ------> 0 p_file_cnt ----------> 5 o_file_cnt ----------> 1 Free Block List: -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Robert DunnMiller Sent: Thursday, May 11, 2006 11:42 AM To: u2-users@listserver.u2ug.org Cc: [EMAIL PROTECTED] Subject: RE: [U2] Unidata logging of BASIC Runtime errors I have seen this happen before when the minimum.modulo is too small (usually the original size of the file). As to the "#" delimiter in the key, that will contribute to poor hashing, but I would recommend the first step to be resizing, and resetting the minimum.modulo to the same size as the resize modulo. HTH --- Regards, Robert -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Baakkonen, Rodney A (Rod) 46K Sent: Thursday, May 11, 2006 09:07 To: 'u2-users@listserver.u2ug.org' Subject: RE: [U2] Unidata logging of BASIC Runtime errors How about some other hashing algorithms like Universe has. I have some files that have compound keys, where the separator is a '#' and the hashing is awful. This is what GROUP.STAT looks like: File = AUTHORIZATION modulo=529314 hash type=0 blocksize=16384 Split/Merge type = KEYDATA Grp# Bytes Records 0 770 1> 1 11583 18>>>>>>>>>>>>>>>>>> 2 0 0 3 12109 22>>>>>>>>>>>>>>>>>>>>>> 4 0 0 5 9647 17>>>>>>>>>>>>>>>>> 6 0 0 7 12215 20>>>>>>>>>>>>>>>>>>>> 8 0 0 9 10338 19>>>>>>>>>>>>>>>>>>> 10 0 0 11 10776 19>>>>>>>>>>>>>>>>>>> 12 0 0 13 10281 18>>>>>>>>>>>>>>>>>> 14 0 0 15 10471 19>>>>>>>>>>>>>>>>>>> 16 0 0 17 12115 22>>>>>>>>>>>>>>>>>>>>>> 18 0 0 19 10812 19>>>>>>>>>>>>>>>>>>> I have experimented with blocksize, KEYONLY, HASHTYPE of 1 and this was the best I could do. I am sure that if I brought this file over to Universe, I could find an algorithm that could handle this compound key a lot better. If you need sample data for files that don't hash well, I have some good examples besides this one. The developers seem to love intelligent keys, and sometimes what they come up with does not hash worth a damn. - Rod -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Wally Terhune Sent: Wednesday, May 10, 2006 10:38 PM To: u2-users@listserver.u2ug.org Subject: Fw: [U2] Unidata logging of BASIC Runtime errors So - we are starting our planning cycle for the next release of UniData now. Is this something that there is a ground-swell of support for? If so - what would it look like? All or nothing switch? UDT.OPTIONS? udtconfig setting? environment variable? All users? Just selected users? New log name - eg unibasic.errlog vs udt.errlog? Some limit to the number logged before shutting it off? (concern about log size and filling disk?) ALL UniBasic run-time errors or warnings logged - regardless of fatality? Just some class of errors? Is so - how would you define the ones you want logged? Collect your thoughts and funnel them thru the U2UG Better and Better committee... That is likely a better way than flooding this list. Just my recommendation. Wally Terhune U2 Support Architect IBM Information Management 4700 South Syracuse Street, Denver, CO 80237 Tel: 303.773.7969 Fax: 303.773.5915 [EMAIL PROTECTED] ----- Forwarded by Wally Terhune/Denver/IBM on 05/10/2006 05:27 PM ----- "Bill Haskett" <[EMAIL PROTECTED] os.net> To Sent by: <u2-users@listserver.u2ug.org> [EMAIL PROTECTED] cc stserver.u2ug.org Subject [U2] Unidata logging of BASIC 05/10/2006 04:07 Runtime errors PM Please respond to [EMAIL PROTECTED] er.u2ug.org Dave: UniData does not have logging for BASIC runtime errors (IBM got back with me on this). There aren't enough requests for this feature to include it in UniData. :-( Bill Haskett ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/ [demime 1.01d removed an attachment of type image/gif which had a name of pic00485.gif] [demime 1.01d removed an attachment of type image/gif which had a name of ecblank.gif] ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/ ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/ ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/ ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/ [demime 1.01d removed an attachment of type image/gif which had a name of pic12872.gif] [demime 1.01d removed an attachment of type image/gif which had a name of ecblank.gif] ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/