If there's any I/O going to the filesystem at all, GPFS has to keep it internally mounted on at least a few nodes such as the token managers and fs manager.
I *believe* that holds true even for remote clusters, in that they still need to reach back to the "local" cluster when operating on the filesystem. I could be wrong about that though. On Sun, Jun 9, 2019, 09:06 Oesterlin, Robert <[email protected]> wrote: > Thanks for the suggestions - as it turns out, it was the **remote** > mounts causing the issues - which surprises me. I wanted to do a “mmchfs” > on the local cluster, to change the default mount point. Why would GPFS > care if it’s remote mounted? > > > > Oh - well… > > > > > > Bob Oesterlin > > Sr Principal Storage Engineer, Nuance > > > > > > > > _______________________________________________ > gpfsug-discuss mailing list > gpfsug-discuss at spectrumscale.org > http://gpfsug.org/mailman/listinfo/gpfsug-discuss >
_______________________________________________ gpfsug-discuss mailing list gpfsug-discuss at spectrumscale.org http://gpfsug.org/mailman/listinfo/gpfsug-discuss
