Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Tuesday 12 August 2008 12:52, you wrote: > Alex wrote: > > Hello experts, > > > > I read that raid software on linux is not cluster aware, so i'm trying to > > find a solution to join together more computers to form a shared file > > system and build a SAN (correctly if i am wrong), avoiding usage of raid > > software. > > > > Let say that I have: > > - N computers (N>8) sharing their volumes (volX, where X=N). Each volX is > > arround 120GB. > > - M servers (M>3) - which are accessing a GFS volume (read/write) > > - Other regular computers which are available if required. > > > > Now, I want: > > - to build somehow a GFS on top of vol1, vol2, ... volN volumes with high > > data availability and without a single point of failure. > > If You want to use GFS, You would need one volX exported via iSCSI or > GNDB to M servers. This exactly what i have now on one of M servers: [root at rhclm ~]# lsscsi [0:0:0:0] disk IET VIRTUAL-DISK 0 /dev/sda [1:0:0:0] disk IET VIRTUAL-DISK 0 /dev/sdb [root at rhclm ~]# here, sda and sdb are block devices imported via iscsi from computer1 and computer2. Question: is possible to group sda and sdb using raid1 software in an array (/dev/md0) and after that, on top of md0 to create a logical volume and run GFS? AFAIK THIS DESIGN IS IMPOSSIBLE because raid software on linux is NOT CLUTER-AWARE? So, just using iscsi (or gnbd) to export volX is NOT ENOUGH because if i am loosing one computerX i am loosing all data on GFS! Right? > For HA you could use DRBD - two volumes vol1, vol2 > created as /dev/drbd0 and exported to M servers. yes, thats why i want to use drbd, to replace raid1 limitations and group computerX two by two. In this scenario, remain just one question: how can i join together /dev/drbd*, or....if this thing is not possible, the question is: how can i join together all 8 volX in order to have: - Fault-tolerance: Failure of a single drive (volX) or server (computerX) should not bring down the GFS! > Your solution isn't good for GFS fs. i am looking to find it, that's why i'm here... can you suggest me one? > You can also use other HA/Cluster fs like: hadoop, gluster, kosmos-fs, > mogile and much more. I read about lustre (from SUN Microsistems). It seems that is well supported on linux (centos5/rhel5) and has support for raid/lvm/iscsi, is scaling well and is easy to extend. Is that correct? Using lustre can i join all volX (exported via iscsi) toghether in one bigger volume (using raid/lvm) and have a SHARED STORAGE? I have one doubt regarding lustre: i saw that they are using EXT3 on top, which is a LOCAL FILE SYSTEM not suitable for SHARED STORAGE (different computers accesing the same volume and write at the same time on it). So, using their patched kernels, ext3 become suitable for SHARED STORAGE? Regards, Alx