Slashdot Mirror
Linux Kernel 2.6.32 LTS Has Reached End of Life
prisoninmate writes: At the end of January we reported the fact that the oldest long-term supported kernel branch, Linux 2.6.32, is about to reached its end of life in February 2016, as announced by Willy Tarreau, who said that there might be another point release in a few weeks if important things need to be fixed. Well, it took a little bit longer than two weeks, and on March 12, he published details about the last maintenance release in the series, Linux kernel 2.6.32.71 LTS, along with the official end of life announcement, recommending users to move to the Linux 3.2 branch.
Would have been helpful if the post had included how old 2.6.32 is.
Last time I wrote a driver (PCI) we were deciding if we wanted to stick with 2.2 or upgrade to the new fangled 2.4 kernel. Last time I dug into the kernel guts (our hardware was really slow for some reason) 2.4 was still considered new.
I'm just a user now using Linux to write software for embedded systems.
That's the last kernel that I as an individual was vaguely able to keep up with. Now it's such a bloated mess that unless you've been part of the development team for decades, you're really not going to be able to maintain an understanding. It seems nearly every project in computing goes this way. Oh well, sic transit and stuff.
According to this chart 2.6.32 had ~9.8 MLoC so I'd say you weren't doing too shabby. In fact if you can keep up with over half the current kernel at ~16.8 MLoC, Linus might be hiring...
Live today, because you never know what tomorrow brings
As someone having to develop complex drivers for 3 years for this particular version of kernel because Cavium would not port their SDK to anything newer until couple years ago, I can honestly say, it was time... for a long time...
If programs would be read like poetry, most programmers would be Vogons.
Oh here we go again! Linus wants more money and obsolete perfectly good computers so they can sell more apps on the Gnu appstore. THis is OUTRAGEOUS! We just upgraded from kernel 2.2 from 2001 last year and Linus ends support already?? Kernel 2.6.32 works just fine and is modern and well supported.
I can't believe users put up with this vandalism and forced obsolence. All the icons are in the wrong spot and my users can't handle change so quickly
http://saveie6.com/
In 2009, three kernels were released, Linux 2.32, FreeBSD 7, and Windows 7. FreeBSD 7 went eol four years later, in 2013. Windows 7 service packs also ended in 2013, Windows 7 mainstream support ended in 2015, Linux 2.32 will support will (somewhat) end in 2016.
So other operating systems have support for 4-6 years, Linux for 7 years.
I know this is in jest but, the reality is that if it is not connected to the internet and it still runs, whatever. Leave it as it is unless you have another reason to upgrade. If you want it on the internet, then it needs to be up to date to deal with all the craploads of malware out there.
"Be particularly skeptical when presented with evidence confirming what you already believe." -
Didn't they backport systemD to that Kernel?
Scientific Linux 6.0 thru 6.6 also use kernel 2.6.32. I'm seeing kernel-2.6.32-573.12.1.el6.x86_64.rpm dated 15-Dec-2015 as the newest SL6 kernel, The Upstream Vendor says they'll be supporting EL6 for 10 or 11 years, so roughly until 2020. Perhaps they'll be backporting changes from newer kernels to 2.6.32?
--- Often in error; never in doubt!
You are confusing the kernel and the system initialization process when the system boots ;-)
Everything I write is lies, read between the lines.
Don't be fooled by the 2.6.32 in RedHat EL6 kernels. That was only the base level. RedHat constantly backports important updates from newer kernels. The latest version is 2.6.32-573.18.1 and still counting (N.B. 573 is somewhat higher than the 71 on the LTS). Mind you I do find the EL7 kernels to be more responsive, so there is always a good argument to upgrade, but no rush 2020 is still a few years away.
It's unlikely the craploads of malware currently in play would be able to target such an old kernel.
2.6.32 had ~9.8 MLoC
Most of which are device drivers.
current kernel at ~16.8 MLoC
Most of which are device drivers, and some more device drivers.
CLI paste? paste.pr0.tips!
Win 7 is in extended support till 2020. it isn't really end of support till it leaves extended support which is all an enterprise requires to have access to hotfixes and security patches. So that is 11 years vs Linux 7
That seems to be a fairly misleading statement
Windows 7 isn't just a "Kernel", it's an Operating System. Ditto for FreeBSD 7. While the kernel may be the core of an OS, userland certainly plays a significant part as well, particular for a desktop OS. For example, on a Win7 64-bit machine, the actual kernel would probably be something like "6.1.7601.17592"
So comparing those three, it would be more fair to use something like a Linux distribution/version from that era, such as
* "Ubuntu Jaunty Jackelope" (EOL Oct 2010)
* "Debian Lenny" (Archived Feb 2011)
* RHEL6 (Production good until 2020, extended life-support not listed yet).
As someone who is kind of a casual Linux system admin, I just make sure to read the Kernel Newbies on each kernel. It really helps me at least kind of keep up on what's going on.
Bingo, RHEL 5+ matches the 10 year support lifecycle for Windows with the same ongoing support past that mark at $Millions per year. That hasn't always been true but as their customer base has grown into the same enterprise markets that Windows has existed in for some time they have had to match the same customer needs/wants/demands. The two support lifecycles even look largely the same with 5-6 years of feature changes followed by 4-5 years of security only patching. Again this matches the general desires of the customer base, they don't want a new OS to grow stale too quickly but in the last few years of the lifecycle they don't want disruptive changes as they are usually focused on preparing the rollout of the new platform, the crossover phase, and the eventual decommissioning of the old systems. Fewer disruptive changes during that period eases the support burden during that period freeing up resources to work on these other projects. I'm currently living that lifecycle as we work on a multi-year project to transition off of Office 2007/Windows 7 and onto Office 2016/Windows 10 (and will be doing the same with the servers once Windows 2016 finally ships this fall). Being able to just quickly test Windows security patches and being relatively confident that they will not break anything frees up admin, QA, and patching resources to focus on new platform.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.