qconf will only work on the internal GE database, not the bootstrap file.
This means something else would have deleted the bootstrap file, which is
pretty concerning. bootstrap(5) indicates that even modifying bootstrap for
a running system isn't supported, let alone deleting the file, so something
outside GE would have had to have done it.
On Tue, Jun 02, 2015 at 11:22:35AM -0500, Dan Hyatt wrote:
> It has been working for a year.
>
> Then today I went to add new execute nodes to the grid, all my grid
> engine clients and the master share a common /opt/sge directory.
>
> And my whole grid went down with the bootstrap message.
>
> I am asking for a restore from the other group, but I need to understand
> maybe what I did, and can I fix it. They can take days to do a restore
> and this is a production system arggg
>
> Thanks,
> Dan
>
> On 06/02/2015 11:18 AM, Skylar Thompson wrote:
> > Rebuilding the bootstrap file is easy, but possibly unnecessary. You should
> > find out why bootstrap no longer exists - did it live there before?
> >
> > On Tue, Jun 02, 2015 at 11:15:32AM -0500, Dan Hyatt wrote:
> >> is it easier to rebuild the bootstrap file?
> >> Or restore it from tape (assuming the other group is backing it up as
> >> requested).
> >>
> >> Dan
> >>
> >> On 06/02/2015 11:05 AM, Skylar Thompson wrote:
> >>> Did your SGE_ROOT and/or SGE_CELL environment variable settings change?
> >>> All
> >>> the GE binaries expect to find the bootstrap file at
> >>> ${SGE_ROOT}/${SGE_CELL}/common. I suspect that the settings changed, and
> >>> your bootstrap file actually lives elsewhere.
> >>>
> >>> On Tue, Jun 02, 2015 at 10:51:58AM -0500, Dan Hyatt wrote:
> >>>> I was trying to add some new exec nodes, and now my qstat and qsub on my
> >>>> master is giving this error.
> >>>>
> >>>> error: fopen("/opt/sge/default/common/bootstrap") failed: No such file
> >>>> or directory
> >>>>
> >>>>
> >>>> When I go to that directory, the bootstrap file does not exist.
> >>>>
> >>>> What did I do and how do I recover?
>
--
-- Skylar Thompson ([email protected])
-- Genome Sciences Department, System Administrator
-- Foege Building S046, (206)-685-7354
-- University of Washington School of Medicine
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
