[petsc-users] When pull the petsc-dev for update:Connection timed out

[Zhenglun (Alan) Wei]

Dear Dr. Brown,
     You're totally right. Thank you so much,

Alan
> It tries to find the executable in PATH, maybe you need mpiexec -n 
> ./app instead of mpiexec -n app.
>
>
> On Mon, Nov 26, 2012 at 1:35 AM, Zhenglun (Alan) Wei 
> <zhenglun.wei at gmail.com <mailto:zhenglun.wei at gmail.com>> wrote:
>
>     Thank you so much, Dr. Brown.
>        Another little question is that:
>            After I update the PETSc, the
>     ${PETSC_ARCH}/conf/bin/mpiexec does not work and shows:
>     [mpiexec at ...edu <mailto:mpiexec at ...edu>] set_default_values
>     (./ui/mpich/utils.c:1541): no executable provided
>     [mpiexec at ...edu <mailto:mpiexec at ...edu>]
>     HYD_uii_mpx_get_parameters (./ui/mpich/utils.c:1745): setting
>     default values failed
>     [mpiexec at ...edu <mailto:mpiexec at ...edu>] main
>     (./ui/mpich/mpiexec.c:154): error parsing parameters
>            Any idea on this?
>
>     thanks,
>     Alan
>
>>     Sounds like you didn't "hg update" and had a bad mergetool or
>>     something. You can use hg revert to discard any local changes.
>>     Since you have many stale files around, you may want to use hg
>>     clean (this will delete all untracked files!) to get back to a
>>     state like if you had a fresh clone.
>>
>>
>>     On Mon, Nov 26, 2012 at 1:25 AM, Zhenglun (Alan) Wei
>>     <zhenglun.wei at gmail.com <mailto:zhenglun.wei at gmail.com>> wrote:
>>
>>
>>>         On Mon, Nov 26, 2012 at 12:59 AM, Zhenglun (Alan) Wei
>>>         <zhenglun.wei at gmail.com <mailto:zhenglun.wei at gmail.com>> wrote:
>>>
>>>             Dear Sean,
>>>                  Thank you so much for your help. Here I have
>>>             another question.
>>>                  After updating the PETSc, I tried to 'make ex45' in
>>>             /src/ksp/ksp/example/tutorial/. It says that
>>>             'petscksp.h' can not be found; therefore, I commented
>>>             this library in ex45.c to see what will happen. It,
>>>             then, shows that the 'petscdmmg.h' can not be found. Any
>>>             idea on this? :)
>>>
>>>
>>>         Since DMMG is gone, baby gone, this means your environment
>>>         (PETSC_DIR) is still pointing to an old version of PETSc.
>>              Since my PETSc was updated, the PETSC_DIR of the new
>>         version of the PETSc should be the same as the old version of
>>         the PETSc, isn't it?
>>              BTW, it works after I re-configure and make the PETSc.
>>         However, I found that it as a lot of dummy things in the
>>         ex45.c, i.e.:
>>         <<<<<<<<<<local
>>
>>         ========
>>
>>         >>>>>>>>>other
>>
>>         =========
>>
>>              It seems that the update changes the original program of
>>         ex45.c by adding fractions of the new code into the old code.
>>         It confused me a lot. Is there any way that I can just
>>         replace the old code by the new code?
>>
>>         thanks,
>>         Alan
>>
>>>
>>>             thanks,
>>>             Alan
>>>
>>>                 On Sun, Nov 25, 2012 at 4:16 PM, Zhenglun (Alan) Wei
>>>                 <zhenglun.wei at gmail.com
>>>                 <mailto:zhenglun.wei at gmail.com>> wrote:
>>>
>>>                     Thank you Sean,
>>>                         I changed the
>>>
>>>
>>>                     $ cd config/buildsystem
>>>                     $ hg showconfig paths.default
>>>                         to
>>>                     https://bitbucket.org/petsc/buildsystem
>>>
>>>                        It gives me 2 extra changes.
>>>
>>>                 Great! Glad to know it's working.
>>>
>>>                        Although, it always giev me the warning of
>>>                     'certificate not verified',
>>>                     it updated lots of files. Should I just ignore
>>>                     this warning?
>>>
>>>                 You could ignore those warnings, yes, but it would
>>>                 probably be better
>>>                 to upgrade your mercurial and see if the host
>>>                 fingerprint option
>>>                 works. Or you could try to use the ssh protocol,
>>>                 assuming your
>>>                 university proxy allows ssh out.
>>>
>>>                      BTW, the reason I used 'https' instead of
>>>                     'http' is that 'https' is shown
>>>                     in the website you gave to me. :) It may needs
>>>                     to be changed.
>>>
>>>                 Well, as I said previously, bitbucket redirects all
>>>                 http traffic to
>>>                 https, so I don't think changing it to just http
>>>                 would really work.
>>>
>>>
>>>
>>
>>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20121125/f990d0c8/attachment.html>