Liane Praza wrote:
> Liane Praza writes:
>
>> Sorry Garrett, for not seeing this sooner.
>>
>> "Garrett D'Amore" writes:
>>
>>> I'm trying to debug a problem with SMF & inetd as part of the removal
>>> of in.tnamed. Basically, I can't seem to figure out why inetd still
>>> spews this message on the first reboot after an upgrade:
>>>
>>> /var/adm/messages.0:May 12 22:45:11 surya-x4200 inetd[28174]: [ID 702911
>>> daemon.error] Failed to update state of instance
>>> svc:/network/tname:default in repository: entity not found
>>> /var/adm/messages.0:May 12 22:45:11 surya-x4200 inetd[28174]: [ID 702911
>>> daemon.error] Failed to update state of instance
>>> svc:/network/tname:default in repository: No such file or directory
>>>
>>> What I can't tell, is why inetd even knows about network/tname. It
>>> should be _gone_.
>>>
>>
>>> I've posted a webrev of what I've figured out so far:
>>>
>>> http://cr.grommit.com/~gdamore/6656092/webrev/
>>>
>> I'm perplexed as well. The code in postinstall should have done it.
>>
>> (The changes to generic_limited_net.xml are for sourcebase cleanliness,
>> not completion of the delete. The whole prophist file is a pre-s10
>> remenant and should/will just be deleted now that upgrades from pre-s10
>> builds are no longer supported.)
>>
>
> I should clarify... pre-*FCS* S10 builds. I'm not talking about S9 or
> S8 upgrades.
>
So you mean Solaris Express builds prior to S10 FC release, right?
Okay well, this is good news. I look forward to seeing that file
removed soon.
>
>> Does one of the inetd upgrade experts think I've missed some inetd magic?
>>
>> I'll keep looking too, but wanted to get the mail out since it seems
>> nobody had responded to you yet. (Sorry, I was on vacation and am
>> trying to catch up.)
>>
>> To clarify, the message only appears during the first boot after upgrade,
>> not a subsequent one?
>>
Correct. I'm wondering if there is some cruft left behind in an SMF
database somewhere that svccfg delete isn't actually deleting.
For the record, I've been directed to submit an RTI, but I need two
reviewers for these changes. Does your review count at least as one of
them?
This is urgent, because its holding up the snv64 respin.
-- Garrett
