Host showing HAVT issues

could be that one path is not registered. Right click on the storage group, properties, press Cntr+Shift+F12, password:messner. Make sure there is a check mark next to every path.

Hi, I’ve managed to connect the “~managements” paths to the actual storage group. But OS still “unknown”, status =”unmanaged” and Connection Status =”HAVT issues”

if you look in Connectivity Status, all paths for that server are logged-in and registered now ?

The registration of the host is managed by the ESXi host agent running on the host. This is supposed to pass to the array the host name, IP address and other information about the host. Make sure that the LAN connection from the host to the array is correct - that the host is using the correct LAN IP on the host to talk to the array. On support.emc.com see Solution emc254967 for a number of different support articles about host agents and connecting to the array. Also, you need certain TCP ports open on the host - 6389-6392.

glen

This host is one out of the 4 hosts we’ve setup for the past few months. Same build, same network configurations as the other 3. The other 3 has no problem. The CX4 is able to recognize their OS and reports no HAVT issues with them.

One other thing to note, besides the CX4, we also have a CX310 attached to this cluster. The CX310 sees this server no problem and it was able to automatically register this server.

The second thing to note is, this server boots from SAN from the CX4. I have a LUN from CX4 presented to this server and the ESXi is installed on it and boots no problem. No issues are reported from VMware side of things.

No additional LUN is assigned to this server. Only the boot partition.

If the CX310 has no problem detecting it, then I must’ve messed something up in the CX4.

screenshot

So this host boots from the CX4 and when it is up and running it also connects to the CX3?

Do the other hosts boot from the CX4?

In the screen cap, this host is shown as manually registered - which makes sense as you must register the host before you can boot it from the array.

glen

correct, after booting from the CX4, there will be other LUNs from the CX4 and CX3 that it connect to.

I've opened up a ticket and the tech suggest that I restart the management server and see if that helps.

I have the same problem on a VNX series and with different host types (Unix/ESXi).

EMC advised to restart the management servers, which did not work, they have now advised that I need to reboot the SPs - I have not done this yet.

did you ever work this out . I'm running into the same problems

See KB Article 92307 - this lists all the articles about HAVT and what is and is not supported as a HAVT configuration in Unisphere or the HAVT Tool (which you can run on the host to check HAVT).

glen

When the host is not having active paths to both the SP you will see this error. Try High availability verification wizard to confirm.

using naviseccli -User xxs -Password -Scope - -h array-hostname.com storagegroup -connecthost -gname -groupnamex -host hostname

this cli will fix the issue. The reason is the GUI doesn't necessary trigger the cli. So using cli to directly deal with array works.