Archived - Instructions for Mitigating Meltdown and Spectre on Enterprise Agents
Last updated
Last updated
As of February 22, 2018, updates for ThousandEyes Enterprise Agents have been made available and tested for the Meltdown and Spectre Version 1 and 2 vulnerabilities. This document outlines recommendations for mitigating vulnerabilities in Enterprise Agents deployed in customer environments. Follow the instructions below for each Enterprise Agent deployment type to apply security updates to your Enterprise Agents.
See describing ThousandEyes' general response to the Meltdown and Spectre vulnerabilities.
Customers who run their own hypervisor software must patch the hypervisor and the host operating system on which the hypervisor runs, in addition to the guest operating system running the Enterprise Agent. Check the information below for your host operating system and hypervisor, or consult your vendor's support resources for updates.
Customers who use cloud-based service providers to run virtual environments will typically be responsible for updating the operating systems of the guest virtual machines.
Download and install patch
Addresses side-channel analysis due to speculative execution
CVE-2017-5753
CVE-2017-5715
Applies to VMware products
ESXI
Workstation / Workstation Pro
Fusion / Fusion Pro
Download and install patch
Addresses Hypervisor-Assisted Guest Remediation for speculative execution issue
CVE-2017-5715
Applies to VMware products
vSphere
ESXI
Workstation / Workstation Pro
Fusion / Fusion Pro
For Agents deployed in Hyper-V environments Windows Server will need to be updated first. Once the Server has been updated Hyper-V's registries will need to be modified to stop VM to VM and VM to host attacks. Refer to instructions linked above.
Windows Server, version 1709 (Server Core Installation)
Windows Server 2016
Windows Server 2012 R2
Windows Server 2012
Not available
Windows Server 2008 R2
Windows Server 2008
Not available
Underlying virtualization infrastructure should be patched for Meltdown and Spectre
Customers are still responsible to update OS kernels for their hosts
Underlying virtualization infrastructure should be patched for Meltdown and Spectre
Customers are still responsible to update OS kernels for their hosts
Underlying virtualization infrastructure should be patched for Meltdown and Spectre
Customers are still responsible to update OS kernels for their hosts
For virtualized operating systems, follow instructions found under Linux Package Enterprise Agents deployed in physical environments.
Here's the quick & easy version:
Ubuntu 14.04 LTS
Ubuntu 14.04 LTS
3.13
Ubuntu 16.04 LTS
4.13 HWE
Ubuntu 16.04 LTS
4.4
*refer to patch notes to check if your kernel version has been patched.
All patched Kernels
RHEL 6.x and CentOS 6.x
Patched Kernel
kernel-2.6.32-696.18.7.el6
RHEL 7.x and CentOS 7.x
Patched Kernel
kernel-3.10.0-693.11.6.el7
You can check your current kernel version with "uname -rs"
If your kernel is not currently patched you can retrieve the latest updates with "yum update". Once run, check that a patched kernel has been loaded with "rpm -qa | grep kernel" - then reboot to apply the update.
After the reboot, run "uname -rs" once again to verify the new kernel has been loaded.
Due to Docker's design, which shares the kernel of the container with the host, there are no upgrades needed within deployed Docker containers. Please refer to appropriate operating system documentation in order to update the host's underlying operating system.
This article will be updated as more updates are made available.
See this link:
See this link:
See this link:
For Ubuntu based systems to upgrade to the latest kernel run the following commands, or follow instructions found at
The latest security updates from Ubuntu should have been applied to your Appliances automatically. Since this change affects the kernel, customers will be required to reboot the Agent. Review the article for more information on the proper way to verify that the kernel has been updated and actions required.
We do not expect a significant performance impact based on testing done in our lab environment. See for information showing details on the nominal performance impact.