Is it what you think or what you saw?there's nothing because
CHR is certainly compatible with ESXi 6 and 6.5(u2).is it CHR not compatible with ESXi 6 ?
well i don't have any problem with memory only cpu and disabled for unknown reason to me !! thanks for your answerI don't know if it's your case, but I've seen CHR stopping when it didn't have enough memory. I originally had VMs with 128MB, but CHR got a little bit more hungry over the time.
that's my real problem only that log the "The CPU has been disabled by the guest operating system" i have so and i don't know why cpu disabled !CHR is certainly compatible with ESXi 6 and 6.5(u2).is it CHR not compatible with ESXi 6 ?
You must surely have some log files purporting to this?
well thanks ,but no not helpingDon't know if this will help but have you read this?
https://support.hpe.com/hpsc/doc/public ... -c00653544
deployed as OVA from mikortik ,is that effect on cpu and may be disable it ?When you say you have just installed CHR, do you mean you’ve used the premade OVA from MikroTik or you have used an ISO to install or how have you done it?
Googleing the problem refers you to the VMware help center and the general consensus is it is caused by a kernel panic in the guest.
So how have you deployed it?
CPU is 32 and ram 8G ,i didn't try 1 CPU cause i have more than 1 Gbps trafficHow many CPU's are you assigning to the CHR installation? What happens if you set just 1?
yes i already paid for ESXi and CHRESXi on the free license will only allow up to 8 vCPUs, have you paid for ESXi? Usually ESXi would not start the VM and tell you to pay for it but maybe you’re getting a weird behaviour?
How many CPUs are you allocating?
well i didn't try for 2 core ,i used to 32 core and get disabled don't know why and seems mikrotik also didn't have a solutions !If you run maybe 2 cores do any get disabled?
If disabling conn tracking will disable NAT, what kind of router is it??Guys, disable conn tracking and problem will gone.
It's been known for a long time. It dies once DDOS kicks in. They not fixed that, but I can confirm disabling conntracks not causing CPU to go offline and machine halt.
A router without NAT is definitely not a home routerIf disabling conn tracking will disable NAT, what kind of router is it??
Hi Alex,Hi,
I recommend taking a look at your vswitch and physical switch architecture. Be clear about your trunks and any spanning tree issues.
Try out the options with promiscuous mode on the vswitches, and esp on the physcial host interfaces.
Make sure to use VMXNET3 interfaces and drivers. Understand clearly about using vlan 4096 in esxi for trunk ports.
I have vsphere 6.5 with 6 CHRs doing BGP, OSPF IP Transit ipv4 and v6 and have NEVER seen this. We have been doing this since 2017.
Our routers are connected via ESXI trunks to external physical switches which we use as port extenders to the CHR, and then connect to Transit providers from there.
And yes, we have conntrack turned on, and we have also had to handle DDOS situations, but never seen the VM crash.
Also the IPArchitechs guys have run some big labs like this with 10g and I am not aware of them seeing the issue (correct me if Im wrong).
Salaams,
Alex