- ipv6calc - lists.deepspace6.net

6rd address anonymization?
by Peter Bieringer 16 Sep '11

16 Sep '11

Hi, according to the RFC the IPv4 address is somewhere stored in the prefix. Is there any common practice, where it is usually stored to get a chance to anonymize? Regards, Peter

1 0

Re: [ipv6calc] [PATCH] 6rd local prefix calculation
by Peter Bieringer 16 Sep '11

16 Sep '11

Hi Raphael, I did now some adjustments, autodetection and checks and also extend the online help. $ ./ipv6calc --action -h Available action types: auto : Automatic selection of action (default) geneui64 : Converts a MAC address to an EUI-64 address conv6to4 : Converts IPv4 address <-> 6to4 IPv6 address (prefix) genprivacyiid : Generates a privacy interface ID out of a given one and a token prefixmac2ipv6 : Generates an IPv6 address out of a prefix and a MAC address anonymize : Anonymize IPv4/IPv6 address without loosing much information 6rd_local_prefix : Calculate the 6rd prefix from given IPv6 prefix & relay prefix and IPv4 Required options: --6rd_prefix ... --6rd_relay_prefix ... Command line: $ ./ipv6calc/ipv6calc --action 6rd_local_prefix --6rd_prefix 2607:fedc:ff40::/43 --6rd_relay_prefix 6.230.0.0/15 6.231.32.33 No input type specified, try autodetection...found type: ipv4addr No output type specified, try autodetection...found type: ipv6addr 2607:fedc:ff52:210::/60 (autodetect messages can be suppressed using -q) Please try latest version from CVS and confirm proper working - thank you! Regards, Peter Am 15.09.2011 20:07, schrieb Raphael Assenat: > Hello Peter, > > It is true that generally, the IPv6 prefix may change > less often. As long as the local IP stays in the same subnet, > we can reasonably expect the relay prefix and relay IP > (not used here) to stay the same. > > However, with my ISP, I receive this informations from DHCP so I > consider all the parameters to be dyanamic anyway. I will always > call ipv6calc with values according to variables within > a script based on the most recently received information. > Therefore, at least for my use case, having the local IP or > the IPv6 prefix as main parameter makes no difference. > > Best regards, > Raphaël Assénat > > On Thu, Sep 15, 2011 at 07:17:39PM +0200, Peter Bieringer wrote: >> Hi Raphael, >> >> patch applied, working. >> >> But I have a question regarding the main input argument: >> >> Your implementation: >> >> ./ipv6calc --action 6rd_local_prefix --in ipv6addr --out ipv6addr >> --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 >> 2607:fedc:ff40::/43 >> 2607:fedc:ff52:210::/60 >> >> But I assume, the IPv6 prefix is a more static one that the local IP. >> >> Because ipv6calc also supports bulk calculations by taking the variable >> argument from stdin, this would be currently the prefix, while I assume, >> the local IP is better. >> >> echo "2607:fedc:ff40::/43" | ./ipv6calc --action 6rd_local_prefix --in >> ipv6addr --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip >> 6.231.32.33 >> >> vs. (implementation must be changed): >> >> echo "6.231.32.33" | ./ipv6calc --action 6rd_local_prefix --in ipv4addr >> --out ipv6addr --6rd_prefix 2607:fedc:ff40::/43 --6rd_relay_prefix >> 6.230.0.0/15 --6rd_local_ip >> >> Opinion? >> >> Peter >> >> At 15.09.2011 05:00, Raphael Assenat wrote: >>> Hello Peter, >>> >>> Here's the patch I omitted in my previous message... >>> >>> Best regards, >>> Raphaël Assénat >>> >>> On Wed, Sep 14, 2011 at 08:39:57PM +0200, Peter Bieringer wrote: >>>> Hi Raphael, >>>> >>>> I can apply your patch but it did not compile: >>>> >>>> librfc5569.c: In function 'librfc5569_calc_6rd_local_prefix': >>>> librfc5569.c:32:26: error: 'struct in6_addr' has no member named 'in6_u' >>>> >>>> Problem is here: >>>> >>>> sixrd_prefix->in6_addr.in6_u.u6_addr8[i>>3] |= 0x80>> (i&0x7); >>>> >>>> Looks like on Fedora 15, the header files are different than on your system. >>>> >>>> Can you rewrite this section using my independend wrapper functions >>>> >>>> ipv6addr_getoctett and ipv6addr_setoctett (looks I have to fix the "tt" >>>> on next release...) instead of deep structure manipulations? >>>> >>>> Please send a patch on top of your previous one. >>>> >>>> Thank you! >>>> Peter >>>> >>>> >>>> At 13.09.2011 19:07, Raphael Assenat wrote: >>>>> Hello, >>>>> >>>>> I recently began playing with the IPv6 service my ISP provides with 6rd. Calculating >>>>> the local prefix manually was tedious so I wrote the attached patch for ipv6calc. >>>>> >>>>> The local prefix is obtained by specifying the 6rd prefix, 6rd relay prefix >>>>> and local IPv4 address to ipv6calc. >>>>> >>>>> The operation is in fact a simple concatenation of the 6rd prefix with part of the >>>>> local IPv4 address (according to the relay prefix length). >>>>> >>>>> Consider the following (fictive) parameters: >>>>> >>>>> 6rd prefix: 2607:fedc:ff40::/43 >>>>> 6rd_relay_prefix: 6.230.0.0/15 >>>>> local ip: 6.231.32.33 >>>>> >>>>> ./ipv6calc --action 6rd_local_prefix --in ipv6addr 2607:fedc:ff40::/43 --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 >>>>> 2607:fedc:ff52:210::/60 >>>>> >>>>> Please review and consider applying. In any case, I think equivalent functionality >>>>> would be a very useful feature to have in ipv6calc. >>>>> >>>>> Best regards, >>>>> Raphaël Assénat >>>>> >>>>

1 0

Re: [ipv6calc] [PATCH] 6rd local prefix calculation
by Peter Bieringer 15 Sep '11

15 Sep '11

Hi Raphael, patch applied, working. But I have a question regarding the main input argument: Your implementation: ./ipv6calc --action 6rd_local_prefix --in ipv6addr --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 2607:fedc:ff40::/43 2607:fedc:ff52:210::/60 But I assume, the IPv6 prefix is a more static one that the local IP. Because ipv6calc also supports bulk calculations by taking the variable argument from stdin, this would be currently the prefix, while I assume, the local IP is better. echo "2607:fedc:ff40::/43" | ./ipv6calc --action 6rd_local_prefix --in ipv6addr --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 vs. (implementation must be changed): echo "6.231.32.33" | ./ipv6calc --action 6rd_local_prefix --in ipv4addr --out ipv6addr --6rd_prefix 2607:fedc:ff40::/43 --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip Opinion? Peter At 15.09.2011 05:00, Raphael Assenat wrote: > Hello Peter, > > Here's the patch I omitted in my previous message... > > Best regards, > Raphaël Assénat > > On Wed, Sep 14, 2011 at 08:39:57PM +0200, Peter Bieringer wrote: >> Hi Raphael, >> >> I can apply your patch but it did not compile: >> >> librfc5569.c: In function 'librfc5569_calc_6rd_local_prefix': >> librfc5569.c:32:26: error: 'struct in6_addr' has no member named 'in6_u' >> >> Problem is here: >> >> sixrd_prefix->in6_addr.in6_u.u6_addr8[i>>3] |= 0x80 >> (i&0x7); >> >> Looks like on Fedora 15, the header files are different than on your system. >> >> Can you rewrite this section using my independend wrapper functions >> >> ipv6addr_getoctett and ipv6addr_setoctett (looks I have to fix the "tt" >> on next release...) instead of deep structure manipulations? >> >> Please send a patch on top of your previous one. >> >> Thank you! >> Peter >> >> >> At 13.09.2011 19:07, Raphael Assenat wrote: >>> Hello, >>> >>> I recently began playing with the IPv6 service my ISP provides with 6rd. Calculating >>> the local prefix manually was tedious so I wrote the attached patch for ipv6calc. >>> >>> The local prefix is obtained by specifying the 6rd prefix, 6rd relay prefix >>> and local IPv4 address to ipv6calc. >>> >>> The operation is in fact a simple concatenation of the 6rd prefix with part of the >>> local IPv4 address (according to the relay prefix length). >>> >>> Consider the following (fictive) parameters: >>> >>> 6rd prefix: 2607:fedc:ff40::/43 >>> 6rd_relay_prefix: 6.230.0.0/15 >>> local ip: 6.231.32.33 >>> >>> ./ipv6calc --action 6rd_local_prefix --in ipv6addr 2607:fedc:ff40::/43 --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 >>> 2607:fedc:ff52:210::/60 >>> >>> Please review and consider applying. In any case, I think equivalent functionality >>> would be a very useful feature to have in ipv6calc. >>> >>> Best regards, >>> Raphaël Assénat >>> >> -- Dr. Peter Bieringer mailto:pb@bieringer.de Heideckstr. phone: +49-89-36109687 D-80637 Muenchen fax: +49-89-36109689 Germany mobile: +49-174-9015046

1 0

Test mesage
by Peter Bieringer 15 Sep '11

15 Sep '11

Hi, sorry, only a test. Regards, Peter

1 0

ipv6calc and ipv6loganon pipe support added/fixed for next release (0.90.0 is on the way)
by Peter Bieringer 15 May '11

15 May '11

Hi, I've commited to CVS latest changes with following major update: - ipv6calc is now able to read input from stdin (got feature request) Example: $ cat /proc/net/if_inet6 | awk '{ print $1 " " $3 }' | ./ipv6calc --ifinet6_to_compressed fe80::38cd:70ff:fe**:****/64 ::1/128 - ipv6loganon is fixed to support pipe also proper, so for on-the-fly anonymization in Apache following can be used: CustomLog "|/usr/local/bin/ipv6loganon -a -f /var/log/httpd/access_log" combined or in conjunction with cronolog: CustomLog "|/usr/bin/ipv6loganon -f |/usr/sbin/cronolog /var/log/httpd/access.log-%Y%m%d" combined Some other features are added like GeoIP IPv6 support and default location for IP2Location database files. Please test current CVS version and report problems, if any. I would probably release final version in a week or so. Regards, Peter

1 0

Re: ipv6calc build system improvements
by Peter Bieringer 06 Jan '09

06 Jan '09

Hi Peter, At 31.12.2008 13:07, Peter Volkov wrote: > Hi. > > В Срд, 31/12/2008 в 11:38 +0100, Peter Bieringer пишет: >>> Second patch is rather big but it modifies rather small number of >>> things: >> I can't apply this patch automagically, patch --dry-run claims a lot. >> Could you please send me a patch against vanilla 0.72.0 - thank you very >> much! > > It was created against vanila ipv6cacl, and I failed to reproduce this > problem. Probably you need to add -p0 option to patch, since by default > it assumes -p1. IOW, Ah, ok, applied and committed to CVS. Please test from CVS, if all is working fine, I would release 0.72.1. BTW: perhaps you have an idea to suppress following message also: config.status: WARNING: 'man/Makefile.in' seems to ignore the --datarootdir setting Regards, Peter

2 2

Re: ipv6calc build system improvements
by Peter Bieringer 31 Dec '08

31 Dec '08

Hi Peter, At 28.12.2008 23:17, Peter Volkov wrote: > Hi Peter. I wrote two patches to improve build system of ipv6calc a bit. > You can find them in attachment. > > First patch optional-geoip makes configure script work as it should in > case user provides --disable-geoip. Ok. > Second patch is rather big but it modifies rather small number of > things: I can't apply this patch automagically, patch --dry-run claims a lot. Could you please send me a patch against vanilla 0.72.0 - thank you very much! Best regards, Peter

1 0

ipv6calcweb and -T
by Peter Bieringer 23 Nov '08

23 Nov '08

Hi, At 19.11.2008 23:32, JC Janos wrote: > make test > > ... > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/ipv6logstats' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calcweb' > # Pseudo random selected > ./test_ipv6calcweb.sh 2001:200:1:ea::1 > Content-type: text/html > > Insecure $ENV{ENV} while running with -T switch at ./ipv6calcweb.cgi line 727. > make[1]: *** [test] Error 2 I can't reproduce this and also I do not understand that Perl claims about it, because I did already the ENV checking and assign only $1 from a regex to the Perl variable. Could you try to figure the reason out and if successful, pls. send me the patch. BTW: I'm using perl-5.10.0-38.fc9.i386 on Fedora 9 Thank you very much! Regards, Peter

1 0

Re: [ipv6calc] Re: ipv6calc 'make update' failure ?
by Peter Bieringer 19 Nov '08

19 Nov '08

JC Janos schrieb: > Hi, > > On Wed, Nov 19, 2008 at 1:43 PM, Peter Bieringer <pb(a)bieringer.de> wrote: >> Pls. try the new 0.72.0 version now, uploaded some minutes before. > > Close, I believe ... > > cd /usr/local/src/ipv6calc-0.72.0/ > > > ./configure \ > --prefix=/usr/local \ > --bindir /usr/local/bin \ > --mandir /usr/local/share/man \ > --enable-geoip > > make update > make > make test > > ... > IPv6/Teredo/RESERVED 0 > IPv6/Teredo/UNKNOWN 0 > > All tests were successfully done! > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/ipv6logstats' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calcweb' > # Pseudo random selected > ./test_ipv6calcweb.sh 2001:200:1:ea::1 > Content-type: text/html > > ?? Use of uninitialized value $bin_ipv6calc in -f at ./ipv6calcweb.cgi line 659. > ?? Error: missing needed program > ?? make[1]: *** [test] Error 1 > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/ipv6calcweb' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calc' > make > make[2]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calc' > ... That's strange, could you please check: cd /usr/local/src/ipv6calc-0.72.0/ipv6calcweb ll ../ipv6calc/ipv6calc This should already exist > sudo make install > > ... > for f in *.8; do \ > install -m 644 $f /usr/local/share/man/man8; \ > done > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/man' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/lib' > make[1]: Nothing to be done for `install'. > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/lib' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/md5' > ?? make[1]: *** No rule to make target `install'. Stop. > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/md5' Will be fixed. Peter

1 1

Re: ipv6calc 'make update' failure ?
by Peter Bieringer 19 Nov '08

19 Nov '08

JC Janos schrieb: > Hi, > > On Wed, Nov 19, 2008 at 12:56 PM, Peter Bieringer <pb(a)bieringer.de> wrote: >>> I am confused, how am I supposed to run 'make install' as root if it >>> invokes 'make test', obviously as root, and you advise to "avoid" >>> doing that? >> Hmm, "make installonly" does the job for you. > > Though ChangeLog says, > > 0.19: Peter Bieringer <pb(a)bieringer.de> > Extend "TODO" list > Print also info on "-?" > Fixes in spec file, see changelog at the end of that file > Add "COPYING" and "INSTALLING" > ----> Add "installonly" on "Makefile" to support changes in spec file > > There's no such target in the Makefile, > > make installonly > make: *** No rule to make target `installonly'. Stop. > > grep install Makefile > install: > ${MAKE} install; \ Just removed, no longer needed, will update the ChangeLog and create a next 0.72.0... Peter

1 0