Thread: [Openpacket-devel] File Naming Conventions
Brought to you by:
crazy_j,
taosecurity
|
From: Mark M. <mas...@gm...> - 2006-07-30 15:16:21
|
Hello - I'm still working on a Rails version of the site. No problem if you end up using David's version. Are you writing any information in the uploaded caps filenames? Source, application, os, date, version information? If I download caps from the site, how do I keep them organized? Take it easy. Mark |
|
From: Tim F. <fu...@cc...> - 2006-07-30 15:49:11
|
Good point - I'd suggest the submitter's userid, a timestamp either of the start of the capture or of the submission, and a short description/title submitted by the user (probably a normalized version of a short free-text title). For the timestamp, the start of capture would be better, but that might be information that the user wants to be anonymized, in which case the submission timestamp might be more appropriate. I don't know about putting too much more information in the filename; if we want it to carry along a lot of meta-info, perhaps we should think about including an associated text .readme or xml .meta_info file (then all of the d/l files would have to be archives), or including fake packets in the trace to carry meta-info (not exactly an elegant solution). I'm just thinking that with all of the meta-info we could possibly include, the filename could start getting a bit silly after a few revs. It might be better to just allow users to look up all the meta-info on the website via filename or md5sum or something, though I'd agree it would be nice to have all the necessary info without having to go back to the website, for various reasons (e.g. lack of connectivity of either the user or the site). -Tim On 7/30/06, Mark Mason <mas...@gm...> wrote: > > Hello - > > I'm still working on a Rails version of the site. > No problem if you end up using David's version. > > Are you writing any information in the uploaded caps filenames? > Source, application, os, date, version information? > > If I download caps from the site, how do I keep them organized? > > Take it easy. > > Mark > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share > your > opinions on IT & business topics through brief surveys -- and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Openpacket-devel mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openpacket-devel > -- Tim Furlong tim...@gm... |
|
From: Jacob H. <ha...@gm...> - 2006-07-31 16:06:09
|
Along the lines of meta information, it might be useful to either build our web app to get statistics about the capture or use tcpdstat (1) and store that as meta information. Jake 1: http://staff.washington.edu/dittrich/talks/core02/tools/tools.html On 7/30/06, Tim Furlong <fu...@cc...> wrote: > > Good point - I'd suggest the submitter's userid, a timestamp either of the > start of the capture or of the submission, and a short description/title > submitted by the user (probably a normalized version of a short free-text > title). For the timestamp, the start of capture would be better, but that > might be information that the user wants to be anonymized, in which case the > submission timestamp might be more appropriate. > > I don't know about putting too much more information in the filename; if we > want it to carry along a lot of meta-info, perhaps we should think about > including an associated text .readme or xml .meta_info file (then all of the > d/l files would have to be archives), or including fake packets in the trace > to carry meta-info (not exactly an elegant solution). I'm just thinking > that with all of the meta-info we could possibly include, the filename could > start getting a bit silly after a few revs. It might be better to just > allow users to look up all the meta-info on the website via filename or > md5sum or something, though I'd agree it would be nice to have all the > necessary info without having to go back to the website, for various reasons > ( e.g. lack of connectivity of either the user or the site). > > -Tim > > > On 7/30/06, Mark Mason <mas...@gm... > wrote: > > Hello - > > > > I'm still working on a Rails version of the site. > > No problem if you end up using David's version. > > > > Are you writing any information in the uploaded caps filenames? > > Source, application, os, date, version information? > > > > If I download caps from the site, how do I keep them organized? > > > > Take it easy. > > > > Mark > > > > > ------------------------------------------------------------------------- > > Take Surveys. Earn Cash. Influence the Future of IT > > Join SourceForge.net's Techsay panel and you'll get the chance to share > your > > opinions on IT & business topics through brief surveys -- and earn cash > > > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > > _______________________________________________ > > Openpacket-devel mailing list > > Ope...@li... > > > https://lists.sourceforge.net/lists/listinfo/openpacket-devel > > > > > > -- > Tim Furlong > tim...@gm... > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys -- and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > > _______________________________________________ > Openpacket-devel mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openpacket-devel > > > |
|
From: Richard B. <tao...@gm...> - 2006-08-01 01:28:12
|
On 7/31/06, Jacob Ham <ha...@gm...> wrote:
> Along the lines of meta information, it might be useful to either
> build our web app to get statistics about the capture or use tcpdstat
> (1) and store that as meta information.
>
> Jake
>
Jake and all,
Great idea -- I suggest capinfos instead of Tcpdstat, since the
protocol breakdown from Tcpdstat isn't very reliable. Capinfos gets
to the heart of what you probably like about Tcpdstat anyway.
Capinfos is packaged with Ethereal/Wireshark.
$ capinfos res_inn.lpc
File name: res_inn.lpc
File type: libpcap (tcpdump, Ethereal, etc.)
Number of packets: 28079
File size: 16661237 bytes
Data size: 16211949 bytes
Capture duration: 2867.890522 seconds
Start time: Thu Mar 30 21:57:25 2006
End time: Thu Mar 30 22:45:13 2006
Data rate: 5652.92 bytes/s
Data rate: 45223.34 bits/s
Average packet size: 577.37 bytes
Speaking of summarization, would statistics like this be of any use?
$ tethereal -nq -z io,phs -r res_inn.lpc
===================================================================
Protocol Hierarchy Statistics
Filter: frame
frame frames:28079 bytes:16211949
eth frames:28079 bytes:16211949
ip frames:27463 bytes:16179237
tcp frames:22417 bytes:15777169
http frames:1558 bytes:1116140
image-gif frames:81 bytes:56366
data-text-lines frames:330 bytes:322423
image-jfif frames:11 bytes:8028
malformed frames:9 bytes:6831
http frames:1 bytes:321
tcp.segments frames:283 bytes:202533
http frames:250 bytes:170943
data-text-lines frames:82 bytes:68786
image-gif frames:68 bytes:47104
media frames:2 bytes:2811
image-jfif frames:21 bytes:19560
ssl frames:33 bytes:31590
ssl frames:426 bytes:252570
malformed frames:19 bytes:11155
pop frames:7915 bytes:11446426
data frames:2 bytes:1755
smtp frames:53 bytes:20255
icmp frames:3984 bytes:221437
udp frames:1053 bytes:180145
dns frames:390 bytes:36121
snmp frames:23 bytes:2737
data frames:46 bytes:2284
bootp frames:109 bytes:37308
nbdgm frames:297 bytes:69841
smb frames:297 bytes:69841
mailslot frames:297 bytes:69841
browser frames:297 bytes:69841
syslog frames:35 bytes:8234
sip frames:20 bytes:9380
nbns frames:124 bytes:12920
http frames:6 bytes:1050
ntp frames:3 bytes:270
igmp frames:9 bytes:486
arp frames:616 bytes:32712
===================================================================
This is less detailed than what the Ethereal GUI provides, but still a
powerful breakdown by protocol.
Richard
|