> > Sorry I cant validate it, my lab is not available now. I took a look at > it and there are some points you must change: > > 1.- lvm-cluster resource definition: the arguments vg_name and lv_name > are mandatory. You must activate a logical volume into the volume group. > As far as i know, you cant build the filesystem on top of the volume > group directly, and I built the script with some checks to avoid that. When I said validate, I meant just check it out. I didn't expect you to run it. Probabily it was not the appropiat word to say what I wanted. (I'm sorry for this poor english.) > 2.- Filesystem resource: you need to mount the filesystem stored into > the logical volume you activate. You need a filesystem to store the VMs, > don't you? Or do you use a "raw device" in your Xen service? No, I don't and can't do that. LVs are passed directly to VMs' configuration files, and Xen takes care of it. Actualy, each LV is seen like a phisical disk by the VM. It's Xen that access and "mounts" it. > I updated your pastebin file. I'm not sure if it will be fully > funcional. > > Maybe you should even nest the resources building your service to be > sure they are started in order. It will depend on your service. Check my > example. Hum... so nesting indicates order. It's good to know. =) > > Yes, but if you have a management console for your VM service you will > need a Virtual IP which "floats" with the service, so you can always > connect to the same IP to check the status of your virtualization > service. You can make that attaching a IP resource to your service. > What are the changes that needs to be done to lvm.conf, as advised in the script code? I have seen that there is already a resource called vm. Maybe, lvm-cluster and vm resouces can do everything I need. I hope that it can. =) -- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster