Clay, I am good with the change and everything else looks good.
Thanks, John On Nov 13, 2010, at 11:54 PM, [email protected] wrote: > Hi John, > Really good catch in check-service! That's bizarre no one had caught it > before -- and what the heck was my logic!?! Anyhow, I've changed line 229 to > be a much more sensible: > elif [[ "$exclude" != "true" && "$exclude" != "false"]]; then > > Otherwise, with regards to ai_sd.py, I agree it should use aimdns(1) going > forward. However, as there's not at this time a need for the AI client to be > multihomed aware, I didn't want to introduce more risk into this fix; I don't > see any problem with moving the client to aimdns(1) though -- it should work > great! > > Thank you, > Clay > > On Sat, 13 Nov 2010, John Fischer wrote: > >> Clay, >> >> You've done a whole bunch of good work!! >> >> A couple of comments are below. >> >> Thanks, >> >> John >> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >> usr/src/cmd/auto-install/ai_sd.py >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >> Still using dns-sd will change to aimdns.py or rearchitect >> in next iteration. You do not need to do anything. We need >> to discuss if we want to wait to re-architect it or make changes >> to it to use aimdns.py. >> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >> usr/src/cmd/installadm/check-server-setup.sh >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >> 229 elif (( ${#exclude} > 5 )) then >> >> Why not simply check if exclude != false and exclude != true ? >> The way it is coded right now if the exclude property value >> exists as long as it is shorter then 6 characters then the >> function would say that it is valid. However, the property >> value could be something other then true or false which would >> still be invalid. >> >> On Nov 13, 2010, at 2:12 PM, [email protected] wrote: >> >>> I caught a merge turd (in auto-install/svc/auto-installer) and a dead >>> comment (in setup-service.sh). New webrevs at: >>> http://cr.opensolaris.org/~clayb/webrev_multihomed/final_round_2/3/ >>> http://cr.opensolaris.org/~clayb/webrev_multihomed/final_round_2/3_diff/ >>> >>> Thank you, >>> Clay >>> >>> On Sat, 13 Nov 2010, [email protected] wrote: >>> >>>> Hi all, >>>> We've finally got consensus on pushing the AI multihomed changes. Below >>>> you can find the full and differential webrevs for the code base. This >>>> code is now integrated with John Fischer's aimdns(1) code too. >>>> >>>> A big thanks to Ethan Quach for helping provide lots of feedback on this >>>> code tonight and John Fischer for working with me on getting aimdns(1) >>>> just so! >>>> >>>> Full-webrev: >>>> http://cr.opensolaris.org/~clayb/webrev_multihomed/final_round_2/2 >>>> >>>> Differential-webrev (from last sent out review): >>>> http://cr.opensolaris.org/~clayb/webrev_multihomed/final_round_2/2_diff >>>> >>>> The scenarios I've retested have been: >>>> -------------------------------------- >>>> install server: >>>> - Verified aimdns(1) record for each enabled subnet >>>> - Verified AI webserver started >>>> - Verified DHCP data printed out >>>> - Network macro for each enabled subnet >>>> - Service macro >>>> >>>> * Unconfigured AI server; DHCP server; post Multihomed image; one network >>>> * Unconfigured AI server; not DHCP server; post Multihomed image; one >>>> network >>>> * AI server with a configured service adding another; DHCP server; post >>>> Multihomed image; one network >>>> * AI server with a configured service adding another; not DHCP server; >>>> post Multihomed image; one network >>>> * Unconfigured AI server; DHCP server; post Multihomed image; three >>>> networks, one with two interfaces >>>> * Unconfigured AI server; not DHCP server; post Multihomed image; three >>>> networks, one with two interfaces >>>> * Unconfigured AI server; DHCP server; pre-Multihomed image; one network >>>> * Unconfigured AI server; DHCP server; pre-Multihomed image; three >>>> networks, one with two interfaces (administratively disabled all but >>>> one) >>>> >>>> * Configured AI server; DHCP server; deleting all services >>>> - Ensured AI webservers stopped >>>> - Ensured AI Apache servers stopped >>>> - Ensured aimdns(1) stopped >>>> - Service in maintenance state >>>> >>>> * Configured AI server; DHCP server; disable a service >>>> - Ensured aimdns(1) record no longer published >>>> - Ensured AI webserver stopped >>>> >>>> * Configured AI server; DHCP server; disable last service >>>> - Ensured AI webservers stopped >>>> - Ensured AI Apache servers stopped >>>> - Ensured aimdns(1) stopped >>>> - Service in maintenance state >>>> >>>> * Configured AI server; DHCP server; enable a service >>>> - Ensured aimdns(1) record published >>>> >>>> * Configured AI server; DHCP server; enable first service >>>> - Ensured AI webserver started >>>> - Ensured AI Apache servers started >>>> - Ensured aimdns(1) started and record published >>>> - Service in online state >>>> >>>> AI client: >>>> ---------- >>>> That the client boots up and passes manifest-locator (e.g. boot image >>>> found; zlib's found; manifest server located) >>>> (All services were tested using automatic DHCP configuration and an X86 >>>> test client) >>>> * Pre-multihomed image with multiple interfaces administratively >>>> disabled to one >>>> * Negative test (boot failed): Pre-multihomed image with multiple >>>> interfaces >>>> * Multihomed image with single interface >>>> * Multihomed image with multihomed interfaces >>>> * Multihomed image with multihomed interfaces but no mDNS record broadcast >>>> (must use "fallback" mechanism which is DHCP BootSrvA for X86) >>>> >>>> Thank you, >>>> Clay >>>> _______________________________________________ >>>> caiman-discuss mailing list >>>> [email protected] >>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >>>> >>> _______________________________________________ >>> caiman-discuss mailing list >>> [email protected] >>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >> >> > _______________________________________________ > caiman-discuss mailing list > [email protected] > http://mail.opensolaris.org/mailman/listinfo/caiman-discuss _______________________________________________ caiman-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
