Arla on OS X 10.2.1: success with klog!

[Thomas Jordan]

kalog is missing from builds 0.35.10pre4 and pre5! klog is still around  
but broken in pre4 (I haven't checked in pre5) Max's solution uses the  
klog from OpenAFS and yields the only way to use Arla on an OSX10.2  
machine.

Interestingly, I have done something similar, used the klog from an  
earlier build of Arla (perhaps pre2?) to get my tokens. It is not  
broken. Perhaps Max's solution and or mine gives the developers some  
place to look to begin fixing the klog in >=pre4.

	t.


On Thursday, September 26, 2002, at 10:27 AM, Todd Cohen wrote:

>
> did you try "kalog" ?
>
> _______________________________________________________________________ 
> ___
> "Yes, but I love gatherings. Isn't it ironic?"
> "Hell hath no fury like a woman scorned for Sega."
>
> http://wckn.clarkson.edu/~cohentl/
>
> On Thu, 26 Sep 2002, Massimo Marino wrote:
>
>> More on that:
>>
>> I have Arla AFS working allright (no ArlaConfiguration GUI though).  
>> The
>> way it suddenly started to work was somewhat twisted.
>>
>> A colleague showed me he had OpenAFS running on OS X 10.2.1 with a  
>> very
>> simply modification: copy kmodstat from 10.1.5 installation of OpenAFS
>> into /usr/bin for OS 10.2.1 . He showed me: worked. OpenAFS is ok for
>> desktop and fixed IP solution. On laptops it is a pain.
>>
>> Anyway: tried on my installation. OpenAFS was working indeed and with
>> 'klog'. It is not aborting as with Arla. Weird. A which klog shows I
>> have it (after OpenAFS installation) on /usr/bin so it is not the one  
>> I
>> was using out of /usr/arla/bin with only Arla installed.
>>
>> Just for the heck of it I restarted the Ti WITHOUT OpenAFS, started
>> arla using 'startarla' and issued a klog: BINGO. Works fine and I got
>> my tickets.
>>
>> Now the ball is in your camp (Arla developers). Arla works: it is its
>> own klog which is apparently broken. Using the klog from OpenAFS fixed
>> the (/usr/arla/bin/)klog abort problem.
>>
>> Cheers
>>
>> 	Massimo
>>
>> On Thursday, September 26, 2002, at 12:47  PM, Massimo Marino wrote:
>>
>>> K,
>>>
>>> I made it work: got tokens and my afs cell up and running with full
>>> access. Running now for an emergency: will post how : will make you
>>> wonder (and maybe laugh!!!)
>>>
>>> Massimo
>>>
>>> On Thursday, September 26, 2002, at 11:08  AM, Eric Knauel wrote:
>>>
>>>>
>>>> Hi,
>>>>
>>>> I just installed kth-krb 1.2 and arla 0.35.10pre5 (with
>>>> --krb4=/usr/athena and --disable-mmap) on my MacOS X 10.2.1 (aka
>>>> darwin 6.1) machine and it seems to work, except that I can not get  
>>>> a
>>>> token:
>>>>
>>>> ,----
>>>> | [knauel at albert ~] /usr/arla/bin/kalog
>>>> | Getting ticket for knauel at informatik.uni-tuebingen.de
>>>> | Password:
>>>> | /usr/arla/bin/kalog: failed inserting tokens for cell
>>>> informatik.uni-tuebingen.de
>>>> `----
>>>>
>>>> Well, I guess I actually got a token:
>>>>
>>>> ,----
>>>> | [knauel at albert ~] /usr/athena/bin/klist
>>>> | Ticket file:    /tmp/tkt5324
>>>> | Principal:      knauel at INFORMATIK.UNI-TUEBINGEN.DE
>>>> |
>>>> |   Issued           Expires          Principal
>>>> | Sep 26 10:51:42  Sep 26 21:31:42
>>>> krbtgt.INFORMATIK.UNI-TUEBINGEN.DE at INFORMATIK.UNI-TUEBINGEN.DE
>>>> | Sep 26 10:51:42  Sep 26 21:31:42  afs at INFORMATIK.UNI-TUEBINGEN.DE
>>>> `----
>>>>
>>>> However arla doesn't seem to notice that:
>>>>
>>>> ,----
>>>> | [knauel at albert ~] /usr/arla/bin/tokens
>>>> |
>>>> | Tokens held by Arla:
>>>> |
>>>> |    --End of list--
>>>> `----
>>>>
>>>> Is kalog the right command to fetch my token? There is also a  
>>>> 'aklog'
>>>> which always quits with an error message ("/usr/arla/bin/aklog:  
>>>> Failed
>>>> getting tokens for cell informatik.uni-tuebingen.de."). I have a
>>>> 'klog' command if I compile arla without '--disable-mmap', which  
>>>> fails
>>>> too.
>>>>
>>>> I'm a bit confused by all these klogs, which one is the right one?  
>>>> ;-)
>>>> How do they differ?
>>>>
>>>> Ciao,
>>>>     Eric
>>>> --
>>>> "Excuse me --- Di Du Du Duuuuh Di Dii --- Huh Weeeheeee" (Albert  
>>>> King)
>>>>
>>>>
>>
>> --
>> Massimo Marino
>> NERSC Division - HPC Department
>> Lawrence Berkeley National Laboratory
>> http://www.slac.stanford.edu/~marino
>> On leave at CERN, CH, EP Division, Atlas experiment
>> phone: (+41) 22 767-1288 fax: (+41) 22 767-8350 Office: 40-3-D16
>> alternate email: marino at slac.stanford.edu, marino at mail.cern.ch,
>> Massimo.Marino at cern.ch
>