Overflowed groups take longer to read, so you want to reduce the number of such 
groups. Is the size of this file roughly constant, or does it just slowly grow?

If it's roughly constant, try setting a MINIMUM.MODULO at around 900,000. It'll 
hopefully break up a lot of your badly overflowed groups. Or change it away 
from a dynamic file to a hash-type that better suits your keys.

If it grows steadily, try REDUCING your split figure - maybe 70%. Again, that 
will increase the number of groups, and hopefully reduce your number of 
overflowed groups.

Lastly, I was going to suggest looking at your block size and large records, 
but I notice you don't have any.

Cheers,
Wol

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of kevinsmith
Sent: 27 November 2007 14:59
To: [email protected]
Subject: [U2] Badly overflowed Dynamic files

I have yet to find someone who can give me a straight answer to this.  I have
a file that is badly overflowed (actually, I have a few files this way).  It
is affecting our batch processing with ridiculous complete times.  What
would be the best way to increase performance with this file.  Do I increase
the split load so that splits occur less frequently?  Any assistance would
be greatly appreciated.

Thanks,
Kevin


The stats of the file:

File name ..................   STOCK
Pathname ...................   STOCK
File type ..................   DYNAMIC
Hashing Algorithm ..........   GENERAL
No. of groups (modulus) ....   777564 current ( minimum 1, 14835 empty,
                                            247100 overflowed, 28403 badly )
Number of records ..........   4770913
Large record size ..........   1628 bytes
Number of large records ....   8
Group size .................   2048 bytes
Load factors ...............   80% (split), 50% (merge) and 80% (actual)
Total size .................   2185496576 bytes
Total size of record data ..   1253392129 bytes
Total size of record IDs ...   32664403 bytes
Unused space ...............   899435948 bytes
Total space for records ....   2185492480 bytes

File name ..................   STOCK
                               Number per group ( total of 777564 groups )
                               Average    Minimum    Maximum     StdDev
Group buffers ..............      1.37          1          4       0.53
Records ....................      6.14          1         26       3.55
Large records ..............      0.00          1          1       0.00
Data bytes .................   1611.95        194       6893     936.26
Record ID bytes ............     42.01          4        179      24.39
Unused bytes ...............   1156.74         12       3560     576.33
Total bytes ................   2810.69       2048       8192       0.00


                               Number per record ( total of 4770913 records
)
                               Average    Minimum    Maximum     StdDev
Data bytes .................    262.72        184       2061      25.79
Record ID bytes ............      6.85          4         15       0.40
Total bytes ................    269.56        188       2076      25.93


--
View this message in context: 
http://www.nabble.com/Badly-overflowed-Dynamic-files-tf4880009.html#a13965441
Sent from the U2 - Users mailing list archive at Nabble.com.
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/

Reply via email to