Re: HA test systemvm can't work. help me!

[guang wu <ki...@gmail.com>]

nobody know reason why is it,? give me some suggest !

2012/10/26 guang wu <ki...@gmail.com>

> thanks everybody,
>
> 1.Your router has a corrupt disk? Can you destroy this router, start a
> new VM for the user? Does the new router still have booting issues?
>
> yes,the disk has corrupted, start a new router VM can normal work. but
> i want to know is this a bug or
> other reason cause the disk corrupt and i want to know how to ensure
> disk data integrity and stability?
>
>
> 2.Have you tried using a redundant virtual router?
>
> i don't know how to add a redundant virtual router, can you give me
> some material about add a redundant virtual router? i use basic network.
>
> 3.i want to know why i can't receive your reply for my email? i see receive for mail-list only?
>
> 2012/10/24 guang wu <ki...@gmail.com>
>
> > hi,everyone:
> >    i 'm very interest in cloudstack,but now i have a trouble when i test
> > HA. i can't solution, but may be you can help me. for example:
> > my environment is that : 1 zone, 1 pod, 1 cluster, 2 host(agent), 1
> > primary storage, 1 second storage, basic network, nfs, second storage wit=
> h
> > management-server at the same machine,
> > primary storage at the another machine lone, primary storage and second
> > storage use nfs. cloudstack version is : 3.0.3,  (host/agent)os : centos
> > 6.2, nfs version is :4, Hypervisor: kvm,
> > install and deploy is success.
> >    everything is ok and i begin try test HA pluck put  reticle that=
>   one
> > of host,  when the host don't contain the systemvm(Virtual Router VM or S=
> econdary
> > Storage VM), the HA can success,
> > otherwise HA fail always, i found the router vm can't running ( work
> > normal ), view the vm console, found the vm disk file failure or pid can'=
> t
> > write or TIP user input run level: or  always stay grub>,  why is that
> > result? nfs can't ensure data integrity and stability or other reason?
> >
> >    i upload some picture,this system vm stay state,
> > help me ,thanks everyone.
> >
> >
>
>

RE: HA test systemvm can't work. help me!

[Somesh Naidu <So...@citrix.com>]

If your host loses connectivity to the storage while a VM is running then it may possibly cause corruption (leave disk in an inconsistent state) both at the virtual disk and file system level.

The former is dependent on how hypervisor handles such disruptions and the latter on the guest OS. This may also be to a certain extent dependent on the storage solution being used.

Cloudstack has nothing to do in this regards.

Regards,
Somesh


-----Original Message-----
From: guang wu [mailto:kingkisskill@gmail.com] 
Sent: Thursday, November 01, 2012 11:42 AM
To: cloudstack-dev
Subject: Re: HA test systemvm can't work. help me!

nobody know reason why is it,? give me some suggest !

2012/10/26 guang wu <ki...@gmail.com>

> thanks everybody,
>
> 1.Your router has a corrupt disk? Can you destroy this router, start a
> new VM for the user? Does the new router still have booting issues?
>
> yes,the disk has corrupted, start a new router VM can normal work. but
> i want to know is this a bug or
> other reason cause the disk corrupt and i want to know how to ensure
> disk data integrity and stability?
>
>
> 2.Have you tried using a redundant virtual router?
>
> i don't know how to add a redundant virtual router, can you give me
> some material about add a redundant virtual router? i use basic network.
>
> 3.i want to know why i can't receive your reply for my email? i see receive for mail-list only?
>
> 2012/10/24 guang wu <ki...@gmail.com>
>
> > hi,everyone:
> >    i 'm very interest in cloudstack,but now i have a trouble when i test
> > HA. i can't solution, but may be you can help me. for example:
> > my environment is that : 1 zone, 1 pod, 1 cluster, 2 host(agent), 1
> > primary storage, 1 second storage, basic network, nfs, second storage wit=
> h
> > management-server at the same machine,
> > primary storage at the another machine lone, primary storage and second
> > storage use nfs. cloudstack version is : 3.0.3,  (host/agent)os : centos
> > 6.2, nfs version is :4, Hypervisor: kvm,
> > install and deploy is success.
> >    everything is ok and i begin try test HA pluck put  reticle that=
>   one
> > of host,  when the host don't contain the systemvm(Virtual Router VM or S=
> econdary
> > Storage VM), the HA can success,
> > otherwise HA fail always, i found the router vm can't running ( work
> > normal ), view the vm console, found the vm disk file failure or pid can'=
> t
> > write or TIP user input run level: or  always stay grub>,  why is that
> > result? nfs can't ensure data integrity and stability or other reason?
> >
> >    i upload some picture,this system vm stay state,
> > help me ,thanks everyone.
> >
> >
>
>