It's actually not a sole ls problem. Any listing of not-replicated directory (with huge files, say movies), by any user program will cause this, and so does the cluster responsivnes very slow. Is there any rough time estimations for 2.1 release? 2009/8/4 Vikas Gorur <vikas@xxxxxxxxxxx>: > > ----- "Anton" <anton.vazir@xxxxxxxxx> wrote: > >> Hello friends, >> >> I have sucessully installed Gluster 2.0.4 in the >> configuration of 3 nodes with 2x bricks per node with >> replication and distribution. While adding an extra brick, >> which had to became a replicated copy of another brick I >> have noticed that gluster is doing replication wile >> accessing a non-replicated data in the rel-time - so >> listing the contents of the non-replicated directory with >> large-size files became unacceptably slow, since replication >> of 10GB data takes time over 1GB network - so the question, >> is there any way (options) to have the given replication in >> the backround, so not-affecting the user work? Or this is >> just is not implemented? > > Work on background self-heal is in progress. It will be > part of the 2.1 release. > > Vikas > -- > Engineer - http://gluster.com/ > >
