Perl log analyser for pproxy stats

[the-mk]

I wish I had a personal proxy so I could use your script
but I think it is not worth it installing one for the last few weeks of OGR...

[Bok]

WTF...

[root@quad dnetproxy]# ./csv5.pl 20080930
This might take a while... please be patient.

Overall OGRp2-25 PProxy stats based on 1 days:

Smallest: 0.21 Gnodes, submitted on 30/09/2008
Largest: 972.41 Gnodes, submitted on 30/09/2008
Average: 54.45 Gnodes
Total: 1,135,757.54 Gnodes
Rate: 1,135,757.54 Gnodes/day

Stubspace 1 stubs: 1
Stubspace 2 stubs: 6
Stubspace 6 stubs: 1,248
Stubspace 7 stubs: 18,171
Stubspace 8 stubs: 1,429
Stubspace 9 stubs: 2

Total stubs crunched: 20,857 (0.00342% of project)

[alpha]

Hehe.. that's me again:

Overall OGRp2-25 PProxy stats for 30/09/2008:

Smallest: 0.94 Gnodes by 192.168.2.6
Largest: 156.68 Gnodes by 192.168.2.1
Average: 44.76 Gnodes
Total: 7117.57 Gnodes

Stubspace 1 stubs: 1
Stubspace 2 stubs: 6
Stubspace 6 stubs: 4
Stubspace 7 stubs: 148

Those are my stats for yesterday. I still don't understand why I'm getting those weird stubs all of a sudden, and it only seems to be me. As long as I keep getting credit for them I'm not going to complain though. I may post over at Ars and see if TheJet can shed any light on it.

[alpha]

The good news is that my script will be incredibly trivial to adapt for OGR-NG. All I need to know is how many stubs there are (and if we're still only doing one level of verification) and a breakdown of stubspaces.

I'll wait for the official stats to come online and see what info they provide before trying to contact someone for the gory details.

Also, I'm tempted to adapt the script to support RC5-72 stats as well, where the user can choose which to print based on a command line option.

[Lani]

just feel free to use the rc5-72 script. i'm currently digging in the ppstats.

[Death]

so what's the progress?

Data shown reflects all blocks received as of 08-Nov-2008 at 23:59 UTC. Current time is 09-Nov-2008 21:34:48.

Status Overview
Total Stubs: 9,967,533

and it looks like there will be no phase2.

[alpha]

Yeah, no phase 2 this time.

The script is ready, I just don't have the details of how the stubspaces are split (either diff length, or stub-length or whatever). Neither TheJet or ertyu could help me with that, nor did anyone else have any ideas in the official IRC channel. Maybe someone else has some ideas? I don't see why it's so difficult, the constraints are obviously known by someone for this page to have been created.

If no-one cares about a stubspace breakdown I can just post the script now, but I'd actually like it myself.

The new script version is fully RC5-72 and OGR-26 compatible and it's such a step forward that I've named the new script "pppla" - Perl proxyper log analyser. It's quite a jump from the previous versions and much more polished.

[alpha]

The new script can be found here. Give it a try!