Endless Lifecycle Support | ~10 min read
Key takeaways
- The free security clock for Debian 11 ends on August 31, 2026, when the volunteer LTS team stops supporting Bullseye. Debian’s own Security Team already stepped back in August 2024.
- An unsupported OS is the loudest finding a vulnerability scanner can return, and it puts PCI DSS, HIPAA, SOC 2, and FedRAMP commitments at risk.
- A two-step upgrade to Debian 12, and then Debian 13, is the real fix. When it cannot land before the deadline, TuxCare Endless Lifecycle Support for Debian 11 keeps Bullseye patched so the upgrade happens on your terms.
Ask three Debian admins when Bullseye really goes end of life, and you may hear three different dates. That is not carelessness. It is a side effect of how Debian supports a release: coverage passes through different hands over its lifetime, and the word “supported” quietly changes meaning at each handoff. For Debian 11, the date that counts is August 31, 2026. That is when the volunteer Long Term Support team stops issuing patches, and the last source of free security updates for Bullseye runs dry.
Nothing dramatic happens at midnight. apt still works, and your services still answer. What changes is that the steady stream of CVE fixes you have leaned on for five years simply stops, and responsibility for every newly disclosed flaw lands on you.
This guide walks through what the deadline actually changes, why an unpatched Debian host is a sharper risk than it looks, and the four routes off Bullseye, including the cases where rushing the upgrade is the wrong call.
How Debian support actually works
A Debian release does not get one support window. It gets a relay. For roughly the first three years, the Debian Security Team handles fixes directly, and this is the “full” support most people picture. When the next stable release ships, the older one is handed to the Long Term Support team, a separate effort staffed by volunteers and sponsor companies that prolongs coverage to about five years from release, but only for a curated subset of packages and architectures. After that, a commercial Extended LTS program from Freexian can extend updates to 10 years, though it sits outside Debian proper.
Bullseye is now in the final free leg of that relay. It shipped on August 14, 2021, left full Security Team support in August 2024, and runs on volunteer LTS until August 31, 2026. Because Debian LTS narrows coverage by both package and architecture — Bullseye LTS is limited to i386, amd64, armhf, and arm64 — some components on a given host may already sit outside the supported set. Installing the debian-security-support package will flag which of your installed packages are no longer covered, so check rather than assume.
Here is where the recent releases sit as of June 2026:
| Release | Released | Debian security support | Debian LTS ends | Status (June 2026) |
| 10 (Buster) | Jul 6, 2019 | Ended Sep 10, 2022 | Ended Jun 30, 2024 | Past end of life |
| 11 (Bullseye) | Aug 14, 2021 | Ended Aug 14, 2024 | Aug 31, 2026 | LTS ends in weeks |
| 12 (Bookworm) | Jun 10, 2023 | Ends Jul 11, 2026 | Jun 30, 2028 | Supported – regular support until Jul 11, then LTS |
| 13 (Trixie) | Aug 9, 2025 | Aug 9, 2028 | Jun 30, 2030 | Supported |
The table also shows why the obvious upgrade target has shifted. Buster (10) is already gone, and Bookworm (12) leaves regular support on July 11, 2026, after which it enters its own LTS phase. That leaves Trixie (13) as the release with the longest clean runway. For a team migrating off Bullseye now, Bookworm may be a necessary stepping stone, but Trixie is the better long-term destination.
Figure 1. Each route off Bullseye buys a different amount of patch coverage. The free options end first.
What end of life means for Debian 11
Because the servers keep running, it is easy to treat the deadline as a non-event. It isn’t. The cost is spread across three areas: security, compliance, and the ecosystem around Bullseye.
Every new CVE becomes your problem
Take security first. Debian’s security fixes are developed in the open, so the moment a patch lands for a supported release, anyone can read the commit, see exactly which lines changed, and infer the bug it closes. On a maintained system, that is harmless, since the fix is already deployed. On Bullseye, that same public commit is a free tip sheet: the vulnerable code is still sitting on your servers, now documented in detail. Turning that knowledge into a working exploit used to take real reverse-engineering time. With AI taking on the grunt work, that time keeps falling, and the safe gap between “patch is public” and “exploit is circulating” keeps narrowing.
Because the operating system sits under everything else, the blast radius is large. An unfixed hole in the kernel, OpenSSL, or another package does not stay contained to one service; it exposes the host and whatever runs on it.
Figure 2. A public fix for a supported release is the blueprint attackers use against the unpatched one.
Auditors flag an unsupported OS
Compliance risk does not wait for an exploit. Most vulnerability scanners read the OS version straight off the host, so a fleet of Bullseye machines starts lighting up as “unsupported operating system” on scans after August 31.
PCI DSS expects known vulnerabilities patched on a defined cadence. SOC 2 reviewers treat EOL components as a control gap. HIPAA and FedRAMP both lean on documented patching and vulnerability management, which gets much harder to defend once the OS underneath stops receiving fixes. A single unsupported distro can scatter findings across an otherwise tidy estate, and cyber-insurers have started writing EOL software directly into their exclusions.
The ecosystem moves on without you
The ecosystem around Bullseye also moves on. Package maintainers and third-party apt repositories drop their Bullseye builds. Official container images are rebuilt on bookworm and trixie. Configuration-management modules, monitoring agents, and language runtimes stop testing against 11. Each of these is a paper cut. Stacked across a couple of quarters, they harden Bullseye in place: every new dependency needs a pin or a workaround, and the upgrade you deferred might get more expensive the longer it waits.
Four ways off Bullseye
There is no single correct answer. The right route depends on how critical the systems are, how much operational load you can handle, and whether a clean upgrade can realistically be completed on time. The flowchart below narrows it down quickly, and the details for each option come after.
Figure 3. A quick way to narrow the decision. Self-maintaining is left off because few teams can staff it.
Upgrade to Debian 12, and then Debian 13
This is the destination every Bullseye host should aim for, and for many systems, it is the right move to start now. The wrinkle that is specific to Debian is that you cannot leapfrog. Supported in-place upgrades go one release at a time, so getting off 11 means two dist-upgrades in sequence, each with its own changed defaults, dropped packages, and third-party repositories to re-point. With Bookworm itself about to enter LTS, aim for Trixie if you want the longest runway. Treat it as a full-blown project: rehearse on a clone, snapshot before each hop, and keep a rollback ready in case a service does not come back the way you expected.
Buy Debian Extended LTS from Freexian
Freexian, a company deeply involved in running Debian LTS, sells a commercial Extended LTS program that covers Bullseye security updates through June 30, 2031. The appeal is simple: you stay on 11 and skip the upgrade entirely. Three details decide whether it fits. Coverage is scoped to a package list, with a fixed base set always supported and anything beyond it quoted per package, while some packages get only limited support or none at all. By Freexian’s own published cost model, the price climbs every six months, gently in the first half of the period and steeply later as other customers migrate away. And the kernel is kept current by moving you onto newer kernel versions backported from later Debian releases, rather than holding you on the kernel you run today. For hosts whose packages are in the supported set and whose owners accept rising bills and occasional kernel jumps, it is a legitimate way to buy time.
Maintain the patches yourself
In theory, you can follow CVEs and backport fixes into your own rebuilds. For an operating system, that means owning security maintenance across the kernel, the C libraries, and dozens of services, on the timeline attackers set rather than the one you would choose. A handful of organizations have the headcount and the C and C++ depth to do this properly. Most do not, and a half-staffed version is more dangerous than admitting you cannot, because it produces the feeling of coverage without the substance.
Extend Bullseye with TuxCare ELS
This is often the most practical option when the upgrade cannot land safely before the deadline. TuxCare’s Endless Lifecycle Support for Debian 11 keeps existing Bullseye installs patched after the LTS cutoff.
TuxCare tracks CVEs, backports the fixes, and ships them as signed packages through apt, covering a broad list of server packages and patching the kernel you already run rather than moving you to a newer version — the reverse of Freexian’s approach above. You can request packages that are not yet on the list. A host connects to the TuxCare repository with a short script, with no reboot and no dist-upgrade, and patches arrive through your existing package workflow.
TuxCare also publishes errata and OVAL data, so OpenSCAP and other OVAL-compatible tooling can still read patch status on those hosts. The price is locked for the full contract term, so the cost does not climb as the release ages.
What you are really buying is timing. ELS splits “stay secure” from “finish the upgrade,” so a fixed calendar date no longer dictates engineering work that should be planned. It earns its place when the upgrade cannot land before August 31, when an auditor needs to see continuous, documented patching, or when a fleet is stable enough that a rushed two-hop upgrade poses the greater risk.
When a fast upgrade is the wrong move
On a slide, “just upgrade” wins every time. On real infrastructure, a few things make it the wrong first move more often than people expect.
The two hops are the hard part. Getting from Bullseye to Trixie means passing through Bookworm first, and each dist-upgrade brings changed defaults, removed packages, and third-party repositories that may need re-pointing or replacing – some may not even have a bullseye-to-bookworm path. That is configuration archaeology, and it does not compress to fit a deadline.
Deadline-driven migrations cost more than they look. An upgrade run against the clock ships broken configs, skips edge cases, and trims testing. The consequences will arrive next quarter in the form of incidents and on-call hours, and they frequently dwarf what a calmer, scheduled upgrade would have cost you, even once you count a year of extended support.
Not every host earns an upgrade this quarter. An appliance, a box running a frozen workload, or a fleet already slated for replacement may not justify a two-hop OS migration right now. Keeping those patched on Bullseye until the planned change is sound engineering, not a dodge.
Stop the clock on Debian 11
August 31, 2026, is fixed. After it, free patches for Bullseye end, and every new CVE accrues to whoever is still running it. Moving to Debian 12 and then Debian 13 is the right destination, and it is worth starting now. For servers that will not make it across in time, or that do not justify the disruption yet, TuxCare’s Endless Lifecycle Support for Debian 11 keeps them patched via apt with no reboots or version jumps.
If you want to map this to your own estate, talk to a TuxCare expert about the hosts you need to carry past the deadline. The goal is to pick your migration date yourself, rather than letting the calendar pick it for you.

