5 Essential Tips for Maximizing Your Experience at Nutanix .NEXT for Bloggers
Stretched Cluster isn't available with Nutanix 4.0.1.1, the latest release. It might be released as part of v4.1. As I do not work for Nutanix, I cannot give any further information.
Since I'm not running production yet, I've decided to reinstall all nodes with the vanilla installation ISO using Phoenix. In addition to the 'clean install hypervisor' option, do I need to choose the 'repair SVM (existing data preserved' option, too?
I get that you need to follow the exact install procedure with Phoenix or Foundation which adds all the specific Nutanix-features (such as ESXi configuration, multiple VIBs, additional drivers, etc.), but in my view, the ESXi installation media is just that, a vanilla install ISO. From a Q&A perspective, I get that Nutanix will only support specific tested and tried installation media. The addition Sylvain makes is that Nutanix won't need or won't test against all the (unneeded) drivers in the Driver Rollup ISO, and that any required additional drivers will be tested, supported and released by Nutanix. With that perspective, the driver set in the Rollup ISO and the driver set released by Nutanix will actually differ, which is why the Driver Rollup ISO is not supported: the end result of installing with a vanilla installation ISO and update with the Nutanix-supported and Nutanix-released driver set will differ materially from installing with the Driver Rollup ISO.
It's still not clear to me :-) So it's not supported to use the Driver Rollup ISO (which contains, say, driver A, B and C), but it is supported to use the vanilla install ISO (which doesn't contain driver A, B and C) and install driver A, B and C manually? Again, the end result would be identical, I believe.
Hi Arthur, Thanks for your reply. I feel however that your reply is ambigious: 1. You say the Driver Rollup ISO is not supported 2. You also say 'update ESXi with patches and drivers after installation' Wouldn't no. 2 result in (more or less) the same situation as no. 1 where I have an ESXi-host installed with the most recent drivers patches? As a side note, I understand the install vs. upgrade argument with the Driver Rollup ISO, but an upgrade is not what I'm talking about; it's a fresh cluster install.
That is smart! That'll allow you to re-use whatever locality the cluster design allows for stretching. Would love to be briefed about it to optimize my current design for this upcoming feature!
I have been digging into the Availabilty Domains feature, and I come to the same conclusion. It simply doesn't have site awereness today and as such, cannot handle split brain scenarios very well.If I have two blocks on one site and a third on a second site, the cluster will go down if the first site fails, since there is only one surviving node. In a split brain scenario, only the first site will survive.Steven, can you tell a bit more about the anticipated process to merge both (site-local) clusters into a single stretched cluster? How will that work, and what will happen to data on those two (site-local) clusters?
Already have an account? Login
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
Sorry, we're still checking this file's contents to make sure it's safe to download. Please try again in a few minutes.
Sorry, our virus scanner detected that this file isn't safe to download.
