On 04/12/2017 09:15 PM, Frank Filz wrote: >> I'd prefer option 1, at this point. Saving config in multiple places is > likely to >> cause more of these issues. If, at some point, we need option 2, we could >> add it then. > > I agree. It should be easy for the Gluster code to use > op_ctx_export_has_option instead of the NFSv4_ACL_SUPPORT macro. > > One thing to consider though is that when this option is changed > dynamically, there are cached attributes which either do or do not have ACLs > depending on which way the change goes. This will cause some unexpected > permission checking and erroneous getattr(ACL) until all attributes have > expired. > > BTW, config_samples/config.txt does document the non-changeability of this > option... > > Note also that the FSAL block in an EXPORT is not processed on update. If > any options in the FSAL block will need to be dynamically updateable, then > option 2, update_export() will be necessary (and the FSAL will have to > manage parsing the FSAL block and processing the changes in a sensible way).
Thank you all for your inputs. For now I have modified FSAL_GLUSTER to read this option value from op_ctx [1]. We will evaluate and update if we need any options in FSAL block to be dynamically changed as well. Thanks, Soumya [1] https://review.gerrithub.io/#/c/356999/ > > Frank > >> On 04/12/2017 08:32 AM, Soumya Koduri wrote: >>> Hi Frank, >>> >>> We ran into an issue with using dynamic export options update using >>> "UpdateExport" dbus signal. This dbus signal invokes >>> update_export_param() which shall update all the related export >>> parameters in memory dynamically. But there are few options which each >>> FSAL may choose to store in its local structure variables (like for >>> eg., for ACL I see FSAL_GLUSTER storing the option value as >>> glfs_export->acl_enabled and FSAL_GPFS has use_acl). These local >>> structure variables seem to be getting updated only while creating the >>> export but do not get changed with dynamic options update. >>> >>> To address this, either >>> * we could modify FSALs to not have a local copy but read these export >>> parameters always from op_ctx->export (or) >>> * have a new FSAL api (maybe update_exports - which is no op by >>> default) to be invoked as part of update_export_param() and FSALs can >>> choose to extend it to update their local structure values. >>> >>> Kindly share your inputs. >>> >>> Thanks, >>> Soumya >>> >>> ---------------------------------------------------------------------- >>> -------- Check out the vibrant tech community on one of the world's >>> most engaging tech sites, Slashdot.org! http://sdm.link/slashdot >>> _______________________________________________ >>> Nfs-ganesha-devel mailing list >>> Nfs-ganesha-devel@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel >>> >> >> >> > ---------------------------------------------------------------------------- > -- >> Check out the vibrant tech community on one of the world's most engaging >> tech sites, Slashdot.org! http://sdm.link/slashdot >> _______________________________________________ >> Nfs-ganesha-devel mailing list >> Nfs-ganesha-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel > > > --- > This email has been checked for viruses by Avast antivirus software. > https://www.avast.com/antivirus > > > ------------------------------------------------------------------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Nfs-ganesha-devel mailing list > Nfs-ganesha-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel > ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Nfs-ganesha-devel mailing list Nfs-ganesha-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel
