Hi Krishna, cut_tree/parse i think if the specfile was not set then cut_tree will exit with an error . Well in changelog avati wrote "now tree is not cut :)", now that is nice to way to know that problem was fixed :D. Regards > Good find Harsha. > > All, > Is this related to the cut_tree/parse problem? Anybody was able > to find the cause of that problem? > > Krishna > > On 2/28/07, Harshavardhana <harsha@xxxxxxxxxxxxx> wrote: >> Hi all, >> I found yesterday that glusterfs and glusterfsd both don't read the >> specfiles when they are in the present directory you need to specify it >> through a path which will start from a '/'. like >> >> glusterfsd -f glusterfs-sever.vol -L DEBUG -l /var/log/glusterfsd.log >> now you do a cat on /var/log/glusterfsd.log i get >> >> [ERROR] FATAL couldn't open the file 'glusterfs-server.vol'. >> >> I am attaching a patch for this fix which was found that 'deamon' >> changes >> the working directory of the present program and sets it to '/'. Coz >> when >> i submitted the above command with NODAEMON option it used to work fine >> for both when you deamonize no specfile is read. >> >> I have tested this patch on Debain 4.0 - GCC 4.1 kernel 2.6.18-3-686 >> SMP. >> >> Regards & Thanks >> >> -- >> Harshavardhana >> "Software gets slower faster as Hardware gets faster" >> _______________________________________________ >> Gluster-devel mailing list >> Gluster-devel@xxxxxxxxxx >> http://lists.nongnu.org/mailman/listinfo/gluster-devel >> >> > > -- Harshavardhana "Software gets slower faster as Hardware gets faster"
