Didn’t hit reply all.
From: Andrew McBeath
Sent: Tuesday, 8 September 2015 10:19 p.m.
To: 'Alexander Neilson'
Subject: RE: [nznog] Appeal to bigger brains
Thanks Alexander, Its an E1000e with vmware tools installed – I have seen quirky behaviour from them before but it seems to have been much better of late. I will try ripping it out and assigning a VMXNET3 adapter.
In response to earlier replies. Kyle -No LACP – single uplink from the blade IO module to a single EX2200
Dylan/Liam – Exactly my first thoughts but 2 of us have been through every bit of kit between the VM and the rest of the network trying to spot what the other missed and are both at a loss. It’s not a complicated network.
Will update tomorrow after I change the NIC.
Appreciate the feedback – Glad to know it strikes others as odd too
Andrew
From: Alexander Neilson [mailto:alexander(a)neilson.net.nz]
Sent: Tuesday, 8 September 2015 10:05 p.m.
To: Andrew McBeath mailto:andrew(a)base-2.co.nz>
Subject: Re: [nznog] Appeal to bigger brains
Which network interface type is on each VM Guest? (or is your issue with the host?) VMXNET3 is recommended however you probably need VMWare tools installed to use it. The E1000 and E1000e interfaces have some weird issues in network operation but not that I have seen as even and odd interfaces.
On your network configuration in VMWare how many underlying interfaces are assigned to each V Switch? can you screenshot the V switches and the individual network config screens under properties.
Sounds like their might be some odd check box or interface involvement that causes this issue.
Regards
Alexander
On 8/09/2015, at 9:53 PM, Andrew McBeath mailto:andrew(a)base-2.co.nz> wrote:
I’m throwing this out there because I’m at a loss to figure out whats happening in our lab. I realise it’s a bit outside the scope of this forum but as the subject line says I’m looking for some bigger brains – Beers are involved.
We have an IBM blade chassis – only 4 blades at present – with a management VLAN 50 (192.168.10.0/24) and a data vlan 100 (10.10.10.10.0/24) hooked up to our prod network (10.10.1.0 and 11.11.1.0 –Yes I know this is BAD but I inherited the network and haven’t been able to change it as yet) via a juniper SRX.
Blades are running VMware 5.5 - no nic teaming as yet.
All blades and AMM module (all on 192.168.10 addresses) can get to all parts of the network
VMWare machine (Only one so far – on 10.10.10.0 network - can ping host, can ping external firewall, CAN ONLY PING ODD OR EVEN NUMBERED IP’s DEPENDING ON WETHER IT HAS AN ODD\EVEN IP) ie if its forth octet is .125 it can ping .1, .3, .5 but if I change it to .126 it can only ping .2, .4, .6
Everything I’ve found online points to NIC teams in certain situations but none apply as the blades only have a single NIC enabled and the VM only has one NIC assigned.
Appreciate any feedback
Cheers,
Andrew McBeath
_______________________________________________
NZNOG mailing list
NZNOG(a)list.waikato.ac.nzmailto:NZNOG(a)list.waikato.ac.nz
http://list.waikato.ac.nz/mailman/listinfo/nznog
