Hi All,

Since upgrading to Ingres 10 we've started picking up some strange errros when running stdump.

For example...
statdump -zn3 -o fred -ulust ldb13 -rarchive_test_profile -amachine_cid
W_OP0976 WARNING: Value for rows 20 and 21 have the same value ().
This will cause E_OP0945 when statistics are reimported.
Consider using -zn to increase precision.

If we look at the file of stats we see:
*** statistics for database ldb13 version: 01000
*** table archive_test_profile rows:256 pages:26 overflow pages:0
*** column machine_cid of type integer (length:4, scale:0, not_nullable)
date:2011_03_15 11:07:15 GMT unique values:11.000
repetition factor:23.273 unique flag:N complete flag:0
domain:0 histogram cells:22 null count:0.000 value length:4
cell: 0 count:0.000 repf:0.000 value: -202963
cell: 1 count:0.023 repf:6.000 value: -202962
cell: 2 count:0.000 repf:0.000 value: -184379
cell: 3 count:0.004 repf:1.000 value: -184378
cell: 4 count:0.000 repf:0.000 value: -86926
cell: 5 count:0.012 repf:3.000 value: -86925
cell: 6 count:0.000 repf:0.000 value: -76561
cell: 7 count:0.047 repf:12.000 value: -76560
cell: 8 count:0.000 repf:0.000 value: -13469
cell: 9 count:0.117 repf:30.000 value: -13468
cell: 10 count:0.000 repf:0.000 value: -5490
cell: 11 count:0.457 repf:117.000 value: -5489
cell: 12 count:0.000 repf:0.000 value: -3008
cell: 13 count:0.059 repf:15.000 value: -3007
cell: 14 count:0.000 repf:0.000 value: -2953
cell: 15 count:0.023 repf:6.000 value: -2952
cell: 16 count:0.000 repf:0.000 value: -88
cell: 17 count:0.188 repf:48.000 value: -87
cell: 18 count:0.000 repf:0.000 value: -6
cell: 19 count:0.031 repf:8.000 value: -5
cell: 20 count:0.000 repf:0.000 value: -1
cell: 21 count:0.039 repf:10.000 value: 0

The values mentioned for rows 20 and 21 look different to me.

Furthermore when I reload these stats there is no error produced at either backend or frontend.

Anyone got anything to offer on this one?

Martin Bowes