Nakivo Backup on separate (direct) lan connection

[jannovakIT]

Hello everyone,

I would like ask you for an advise.

This is only about my home lab, so just simple one esxi host without vcenter.

I am also using Nakivo as backup solution for my lab thats running on QNAP.

So far they were both on same subnet running without issues.

This week I added a 10GB connection to both devices and have just direct cable connection between them.

From esxi and VMs on top of it I can reach the QNAP without problem (NFS, SMB, SSH).

Also from QNAP itself I can SSH to the ESXi host.

All this using the 10GB connection.

But when I add this host with a new IP to Nakivo It all behaves very weird.

I know Nakivo is using ports 902 and 443, but nothing is being blocked on esxi side.

The only message I receive in hostd logs on esxi is

Accepted password for user nakivo from 192.168.168.5

2019-04-09T20:15:32.268Z info hostd[2098603] [Originator@6876 sub=Vimsvc.ha-eventmgr opID=6eee63d1] Event 1208 : User nakivo@192.168.168.5 logged in as Java/1.8.0_181

2019-04-09T20:15:33.068Z info hostd[2098603] [Originator@6876 sub=Vimsvc.ha-eventmgr opID=6eee6459 user=nakivo] Event 1209 : User nakivo@192.168.168.5 logged out (login time: Tuesday, 09 April, 2019 08:12:27 PM, number of API invocations: 0, user agent: Java/1.8.0_181)

2019-04-09T20:15:36.182Z info hostd[2099204] [Originator@6876 sub=Solo.Vmomi opID=6eee636e user=nakivo] Activation [N5Vmomi10ActivationE:0x00000098031a9820] : Invoke done [retrieveContents] on [vmodl.query.PropertyCollector:ha-property-collector]

2019-04-09T20:15:36.182Z verbose hostd[2099204] [Originator@6876 sub=Solo.Vmomi opID=6eee636e user=nakivo] Arg specSet:

--> (vmodl.query.PropertyCollector.FilterSpec) [

-->    (vmodl.query.PropertyCollector.FilterSpec) {

-->       propSet = (vmodl.query.PropertyCollector.PropertySpec) [

-->          (vmodl.query.PropertyCollector.PropertySpec) {

-->             type = "vim.HostSystem",

-->             all = false,

-->             pathSet = (vmodl.PropertyPath) [

-->                "config"

-->             ]

-->          }

-->       ],

-->       objectSet = (vmodl.query.PropertyCollector.ObjectSpec) [

-->          (vmodl.query.PropertyCollector.ObjectSpec) {

-->             obj = 'vim.HostSystem:ha-host',

-->             skip = false,

-->             selectSet = <unset>

-->          }

-->       ],

-->       reportMissingObjectsInResults = <unset>

-->    }

--> ]

2019-04-09T20:15:36.182Z info hostd[2099204] [Originator@6876 sub=Solo.Vmomi opID=6eee636e user=nakivo] Throw vmodl.fault.SystemError

2019-04-09T20:15:36.182Z info hostd[2099204] [Originator@6876 sub=Solo.Vmomi opID=6eee636e user=nakivo] Result:

--> (vmodl.fault.SystemError) {

-->    faultCause = (vmodl.MethodFault) null,

-->    faultMessage = <unset>,

-->    reason = "System exception while trasmitting HTTP Response:

--> error id = 32

--> "

-->    msg = ""

--> }

2019-04-09T20:15:36.182Z error hostd[2099204] [Originator@6876 sub=Default opID=6eee636e user=nakivo] Failed to send response to the client: Data already transmitted; cannot reset.

I am sure there is no blocking, network is setup correctly.

The thing I am not sure about is how in general esxi works with 2 management interfaces.

I would be very glad if you have any suggestions.

BR,

Jan

[cyberpaul]

Čau,

not sure how your networking is set. I believe if you have two management interfaces, you should use different subnets on each one. Otherwise the ESXi host might get confused about where to send the data.

So if you use 192.168.168.0/24 for the 1 Gb network, use something different like 192.168.169.0/24 for the 10 Gb network. Then, obviously, reconfigure Nakivo to use the new address for host connection.

If you already use different subnets, I suggest checking network masks on both ends.

Pavel

[aaron009]

If I were I would have asked such a question on their forum.