On 2011-08-29, at 10:33 PM, Pasi Kärkkäinen <pasik@xxxxxx> wrote: 
 
 On Mon, Aug 29, 2011 at 05:18:21PM +0800, 350608693 wrote:   hello,everybody,
     i recently study on the remus,and install xen-4.2-unstable on cetos
     5.5(x86_64),and the kernel version is 2.6.32.40.
  There's a newer kernel already available for dom0, but dunno if that'll help.
  
 Did you try the drbd backend for Remus?http://wiki.xensource.com/xenwiki/RemusI added CC to Shriram, maybe he can help..-- Pasi   but i met a lot of problem,such as:
     1)when using the configure "disk = [
     'tap2:remus:backup:9000|aio:imgpath,xvda,w' ]",after a few minutes,the vm
     looks just as dead and can NOT log in;
    Are you sure that the problem is caused by the disk configuration ? Is imgpath same on both machines?  More importantly, make sure that the disk is not on a shared storage    2)i install a centos hvm guest,and run the command "remus <domid>",error
     message is as follow:
     Disk is not replicated: tap:aio:/home/xen/tpl/hvm/base.img,xvda,w
  
 This is a harmless error This is the actual issue. You ran Remus without specifying a backup host. You have enabled netbuffering but I don't know if you installed appropriate modules.  sch_plug  IFB  other netfilter modules.  Checkout the wiki page that Pasi pointed to.    3)i install a centos pv guest,the configure file is as follows:
     name = "mailserver"
     memory = "512"
     disk = [ 'tap2:aio:/var/lib/xen/images/vmdisk.img,xvda,w', ]
     #disk = [
     'tap2:remus:192.168.226.201:9000|aio:/var/lib/xen/images/vmdisk.img,xvda,w',
     ]
     #disk = [
     'tap2:remus:server02:9000|aio:/var/lib/xen/images/vmdisk.img,xvda,w' ]
     #vif = [ 'bridge=xenbr0', ]
     #vif = [ "type=ioemu,bridge=eth0, mac=00:16:3e:38:9a:ff" ]
     #vif=['']
     bootloader="/usr/bin/pygrub"
     vcpus=1
     #
     #
     
     
  
 
    it can boot properly.but when run the command "remus <domid>",the error
     message is as follows:
     Disk is not replicated: tap:aio:/var/lib/xen/images/vmdisk.img,xvda,w
     WARNING: suspend event channel unavailable, falling back to slow xenstore
     signalling
     PROF: suspending at 1314608468.806815
     PROF: resumed at 1314608468.828319
     PROF: flushed memory at 1314608468.829534
     PROF: suspending at 1314608469.022028
     domain 9 not shut down
     xc: error: Suspend request failed: Internal error
     xc: error: Domain appears not to have suspended: Internal error
     PROF: resumed at 1314608469.035179
  
 I am not sure what kind of kernel is running inside centos. It seems to support xenstore based suspend but something weird is happening. 
 
 
 
    i really do not know what to do next and how to address this problem.could
     anybody tell me how to address them?
     What's more,could anybody tell me how to install xen that supports remus
     properly and how to install PV/HVM domains and the configure file in
     details ?
  
 
 
 Try this wiki (one of Remus users created it) 
 
 Also, in future, if you want to run remus on one system without disk replication and network buffering, try remus -i <interval> --no-net --blackhole domname  |