On Tue, Apr 21, 2015 at 06:07:00PM +0530, Atin Mukherjee wrote: > > > On 04/21/2015 05:47 PM, Avra Sengupta wrote: > > Hi, > > > > Today whenever a gluster command fails the ret code is always 1. > > Irrespective of the failure. I have sent the following patch which takes > > a first step towards bringing some order to this chaos. > Could you explain a bit more about the current problem if we do not have > separate ret code? Different return values for a command is quite common. One of the major examples is documented in 'man 8 mount' in the 'RETURN CODES' chapter. Scripts and other applications can then use the return code to inform the users/admins and act in a more suitable way. Cheers, Niels > > ~Atin > > http://review.gluster.org/10313 > > > > It addresses the following scenarios: > > > > 1. Invalid Syntax: I have incorporated this for the snapshot commands > > only for now. In case of an invalid syntax the retcode will be 2. > > 2. Another transaction in progress errors: In case the command fails to > > acquire a lock, because another transaction is in progress the retcode > > will be 3. > > > > If the approach is fine, and it's fine to go with such incremental > > retcodes, then once this patch gets accepted, I will send further > > patches on top of this one. I would also urge other component owners to > > block retcodes for their components in "glusterd-errno.h", and start > > doing the same. > > > > Regards, > > Avra > > _______________________________________________ > > Gluster-devel mailing list > > Gluster-devel@xxxxxxxxxxx > > http://www.gluster.org/mailman/listinfo/gluster-devel > > -- > ~Atin > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxxx > http://www.gluster.org/mailman/listinfo/gluster-devel
Attachment:
pgph2LXj1mCV9.pgp
Description: PGP signature
_______________________________________________ Gluster-devel mailing list Gluster-devel@xxxxxxxxxxx http://www.gluster.org/mailman/listinfo/gluster-devel
