Unscheduled Maintenance on vServer Hosts - CVE-2026-99999 (critical)
- Type: Service
- Category: vServer
- Start: 2026-10-19T09:00:00+0000
- End: 2026-10-19T18:00:00+0000 (expected)
For security reasons, we need to perform maintenance on our vServer hosts on short notice, which may result in brief, isolated periods of downtime.
Our usual procedure
For scheduled maintenance, we typically use a rolling procedure: The hosts are migrated one by one - that is, all running vServers are first moved to another host so that the host undergoing maintenance is no longer processing active traffic at the time of the reboot. This ensures service reliability, and our customers generally don’t notice anything.
Why we’re taking a different approach in this case
For CVE-2026-99999, this approach is insufficient. The vulnerability is so severe that we must patch and reboot all affected hosts as quickly as possible to minimize the attack window. This means that some vServer hosts may experience a brief interruption (downtime) during the kernel update and subsequent reboot.
Background
A highly critical (CVSS 9.3 out of 10) vulnerability has been discovered in the IPv4 stack of the Linux kernel that could potentially allow code execution all the way up to the host system itself:
Vulnerability CVE-2026-99999 (https://nvd.nist.gov/vuln/detail/cve-2026-99999)
Cause: Faulty checksum metadata in a network packet is not sufficiently validated.
Impact: This can lead to memory corruption, which could potentially allow arbitrary code execution, unauthorized data access, or a system crash.
Remedy: Update the system kernel to a version that includes the appropriate security fix.
We ask for your understanding that, in this case, we are deliberately deviating from our usual downtime-free procedure - the security of your systems is our top priority!
Our usual procedure
For scheduled maintenance, we typically use a rolling procedure: The hosts are migrated one by one - that is, all running vServers are first moved to another host so that the host undergoing maintenance is no longer processing active traffic at the time of the reboot. This ensures service reliability, and our customers generally don’t notice anything.
Why we’re taking a different approach in this case
For CVE-2026-99999, this approach is insufficient. The vulnerability is so severe that we must patch and reboot all affected hosts as quickly as possible to minimize the attack window. This means that some vServer hosts may experience a brief interruption (downtime) during the kernel update and subsequent reboot.
Background
A highly critical (CVSS 9.3 out of 10) vulnerability has been discovered in the IPv4 stack of the Linux kernel that could potentially allow code execution all the way up to the host system itself:
Vulnerability CVE-2026-99999 (https://nvd.nist.gov/vuln/detail/cve-2026-99999)
Cause: Faulty checksum metadata in a network packet is not sufficiently validated.
Impact: This can lead to memory corruption, which could potentially allow arbitrary code execution, unauthorized data access, or a system crash.
Remedy: Update the system kernel to a version that includes the appropriate security fix.
We ask for your understanding that, in this case, we are deliberately deviating from our usual downtime-free procedure - the security of your systems is our top priority!
Update: Emergency maintenance successfully completed
The emergency maintenance has been successfully completed.
Contrary to our initial assessment, we were able to completely avoid downtime for nearly 60% of the vServers. The remaining vServers were unavailable for a few minutes during the necessary reboot of the host system.
Contrary to our initial assessment, we were able to completely avoid downtime for nearly 60% of the vServers. The remaining vServers were unavailable for a few minutes during the necessary reboot of the host system.