Revision history for C and Perl extension of nfdump.
1.06 2014-12-06
- perl: build system update
1.05 2014-12-05
- c: build system update
- c: added --enable-threads for thread support; disabled by default
- c: flex, bison is not needed for building
- c: adjusted to nfdump 1.6.13
1.04 2014-11-17
- perl: added example examples/example2.pl
- perl: added test 08-aggreg-sort.t
- perl: aggregation and sorting added into
Net::NfDump (Aggreg => 1 and SortBy => <field> attributes)
- c: lnf_field_t structure extended with default sort and
aggregation order
- c: added example lnf_ex04_threads.c
- c: threads support added for paralel processing flow
records in aggregation. Code not released for public,
the public version works properly only in sigle
thread enviroment.
- c: added example lnf_ex03_aggreg.c
- c: added lnf_mem object with aggregation and sorting capability
- c: lnf_rec_fget, lnf_rec_fset incerased prtformance (30%)
Thanks to Vasek Bartos.
1.03 2014-06-23
- c: headers includes in the examples
- c: code preparation for aggregation and sorting
- perl: test for pure read withour record processing (-G)
1.02 2014-06-18
- c: added structure lnf_brec1_t containing most commonly used records
- c: examples updated to use lnf_brec1_t
- c: examples updated to be usable for benchmarking
- perl: updated build system (more resistant)
- perl: benchmarking of c code added as the part of perl testing
1.01 2014-06-15
- source code strcitly separated into autonomous
package (libnf-xxx, universal C API) and perl
Net::NfDump module that is placed on top of the C API
- resedigned build system
- prepared for new features (aggregation, sorting, ...)
- changes inspired by Vasek Bartos and his libnfdump code
0.14 2014-04-24
- examples updated to use last version of Net::IP::LPM
0.13 2014-04-20
- adjusted to nfdump 1.6.12
- support for new record type introdouced by nfdump 1.6.12
- transparently reads record types from previous nfdump
- created files are always with the new record type
- !! files created by this version mus be read by nfdump >= 1.6.12
- support for new fields (ingress_vrfid, egress_vrfid, event_flag
block_start, block_end, block_step, block_size)
- flow_start item changed to event_time
- fw_event removed (use fw_xevent instead)
- doc extended with info about aggregation
0.11 2013-11-26
- adjusted to nfdump 1.6.11
0.10 2013-09-12
- decreased test length
0.09 2013-08-13
- adjusted to nfdump 1.6.10p1
0.08 2013-06-06
- removed bug in TimeWindowStart, thanks to Pavel Sejnoha
0.07 2013-05-21
- removed bug in examples, thanks to James Deaton
- updated to nfdump 1.6.10
0.05 2013-05-03
- language corrections
0.05 2013-03-16
- silently skip known (but not supported) blocks
- example added (update AS info)
0.04 2013-03-03
- the first stable version
0.03_02 2013-03-03
- the version of nfdump updated to 1.6.9
0.03_01 2013-02-28
- NEL, NSEL support added
- a few bugs removed
0.02_05 2013-02-26
- example1.pl added
- documenation updated
0.02_04 2013-02-24
- a few bugs removed
- clonerow added
- InputFiles and Fields can by handled as a reference to list or comma separated list
0.02_01 2013-02-11
- hasharray functions added
- performance optimalisation (4x faster)
0.01_06 2013-01-20
- nfdump source codes added into distribution
- fixes for libnf on OpenBSD apllied
0.01_04 2013-01-15
- extra compilation of nfdump utility
0.01_04 2013-01-14
- removed dependency on lex/flex during compilation
0.01_03 2013-01-12
- fixes for multiplaform compilation
0.01_02 2013-01-10
- documentation update
0.01_01 2012-11-30
- initial version