Single Point of failure iSCSI

Johnjones

Hi all,

I've got StarWinds iSCSI software running on a VM within ESXi host. A couple of other virtual machines are running from that shared storage. If the storage VM goes offline, I lose everything. Any suggestions on how to combat this?

Thanks.

Essendon

The host is a SPOF too! Unless you have some kind of replication going on, I dont think you can do much about it. I believe the free version they have offers async (maybe sync too) replication to a 2nd cluster, check that out. Other than that, SRM? Again the host is a SPOF on its own.

jibbajabba

Like Ess says, your host is single point of failure as well. We used to say, every 9 after the comma of uptime adds a 0 to the price.

If you are worried about data loss, think about backups. There is a free version of Veeam for example. You could add large disk / raided disk to your host, create a Veeam VM and backup your VMs daily / weekly to that datastore - or get a NAS.

If you do just have one host, then you probably don't need shared storage to begin with, unless you run nested clusters ?

Anyway, like I say - if worried - make backups ..

Johnjones

Thanks for the replies. I'm running one host with 2 nested esxi. I only have one host going because of concerns about energy consumption. It appears I may have to add another though.

dave330i

Is this a lab or production? If it's a lab, why do you care about SPoF? If it's production, why are you running nested?

Johnjones

Of course it's lab. I was curious what people do in these situations. I can't imagine anyone running nested in prod.

jibbajabba

Rather than investing in a second host, invest in a NAS and backup

jibbajabba

Johnjones wrote: »

Of course it's lab. I was curious what people do in these situations. I can't imagine anyone running nested in prod.

From the support and license perspective probably nobody does. But I know a few companies who do. Granted, just DEV / QA and the sorts, but they do