I have an application with is binary-only, does its job well, and is only available for either libc5 (!) or early early glibc2.0 (!!). It has been running on a Red Hat Linux 5.2 (NOT RHEL; RHL) server for a really long time, and it honestly does its job and it's not easily replaced by an open-source solution at the moment.
So, I need to do one of the following things: 1.) Run Red Hat Linux 5.2 (or similar vintage) on KVM on CentOS 7; 2.) Build libc5 (!) for CentOS 7 and run it on C7 (if that's even possible).
The latest version of libc5 I know of that was shipped by Red Hat is in RHL 6.2, libc-5.3.12. (There is a 5.4, but not sure of stability or compatibility).
So, I have RHL 5.2 (Apollo) running on KVM on C7, but I'm not getting any networking working. The only two non-virtio NICs are e1000 and rtl8139. As I recall kernel 2.0.36 and rtl8139 were pretty iffy together, and e1000 drivers are not likely available (I couldn't find any). The host's bridge sees the MAC of the guest, but the guest isn't seeing any traffic.
So, has anyone tried doing anything this old with modern kvm (I'm using the qemu-ev packages, incidentally)?
I've successfully set up the bridging; a CentOS 7 VM on the same host has full connectivity. So it's something about the rtl8139 and the 2.0.36 kernel. What is the oldest distribution you've done on KVM on C7?
On 04/21/2017 12:11 PM, Lamar Owen wrote:
... The latest version of libc5 I know of that was shipped by Red Hat is in RHL 6.2, libc-5.3.12. (There is a 5.4, but not sure of stability or compatibility). ... I've successfully set up the bridging; a CentOS 7 VM on the same host has full connectivity. So it's something about the rtl8139 and the 2.0.36 kernel. What is the oldest distribution you've done on KVM on C7?
Ok, so I've progressed somewhat on this. Here's what I've so far found: 1.) RHL 5.2 installs fine, but is a no-go, since it has no networking. 2.) RHL 6.2 installs fine, but is a no-go, with lots and lots of 'hda: lost interrupt' errors and eventually the guest is paused and won't proceed past a certain point in the boot. Of all the days for the Fedora infrastructure to go completely down, while I'm looking for archived RHL (!!!). For a while I couldn't even get to archive.org. Trying to get to bugzilla.redhat.com was impossible, and even archived mailing list posts were not available talking about the lost interrupt issue (google's cache will only help you so far). It was a pretty odd sight, seeing a 'www.redhat.com is down but will be back soon' webpage come up. 3.) vault.centos.org to the rescue; CentOS 2.1 is essentially an RHL 7.x-derived distribution, and so I grabbed the two ISOs from a vault mirror. The GUI install crashes out completely, but a text install worked swimmingly. Since the app I need does have a glibc2.1 version, and since I already had it downloaded from back in 1998 or so, using the RHL 6.2 compat-glibc2.1 packages should work. Installed, booted, and got networking. The C2.1 guest seems stable, and I'm pulling the updates repo from vault.centos.org and will set up local repos for yum (yum version 1!) and get it as fully updated as C2.1 can be (this isn't exposed to the internet). Once updated and the compat-glibc2.1 installed will try out the app in a testing mode, then pull over the database (postgresql 6.5, but I have a good dump I can restore into the C2.1 postgresql) and bring it back live.
Now, there is one other option that I found this morning: the LinuxTech repo for CentOS 6 has a libc5 for CentOS 6. I'm setting up a testing C6 i386 guest, and will try out the LinuxTech libc5. If that works that will be the route I go, since C6 is still updated. I'm not going to abandon the C2.1 setup yet, though, since the app I'm using may need some other libs. We'll see.
So, to answer my own question, CentOS 2.1 is the oldest thing I have thus far successfully run with full server functionality in C7's qemu-kvm-ev.
Am 22.04.2017 um 15:42 schrieb Lamar Owen lowen@pari.edu:
On 04/21/2017 12:11 PM, Lamar Owen wrote:
... The latest version of libc5 I know of that was shipped by Red Hat is in RHL 6.2, libc-5.3.12. (There is a 5.4, but not sure of stability or compatibility). ... I've successfully set up the bridging; a CentOS 7 VM on the same host has full connectivity. So it's something about the rtl8139 and the 2.0.36 kernel. What is the oldest distribution you've done on KVM on C7?
Ok, so I've progressed somewhat on this. Here's what I've so far found:
Silly question: isn’t there something like a „compat-CentOS5“-package one can install and that contains all the base-libraries for CentOS 5?
I run a FreeBSD 6 (32 bit) binary on a 64 bit FreeBSD 11 VM (because the source, if we had it, would most likely not compile with whatever LLVM ships with 11 …). FreeBSD offers „compat“ packages down to version 4. These are libraries that install into /usr/lib/compat.
Obviously, the system is designed for this - but why is nobody doing this for Linux?
I just looked it up: FreeBSD 4.0 was released over 17 years ago, around the same time as RHL 6.2…
On 04/22/2017 11:24 AM, Rainer Duffner wrote:
Obviously, the system is designed for this - but why is nobody doing this for Linux?
The time commitment necessary to keep up with the sometimes dramatic changes in the typical Linux distribution is very large; I would guess that's the reason it's not typically done.
The LinuxTech EL6 libc5 is only possible because Mandriva 2011 contained it. I've not tested that piece yet, but plan to a bit later.
I just looked it up: FreeBSD 4.0 was released over 17 years ago, around the same time as RHL 6.2…
Yep, and the binaries I have are older than that. The particular setup being used actually dates from 1997, almost exactly 20 years ago. The version was upgraded a few times afterwards; the timestamps on the original distribution archive for the glibc 2.1 version are all from May 12, 1999, but the original version was from fall of 1996.
I'm glad I held those two tarballs back all these years.
In any case, CentOS 2.1 plus the RHL 6.2 compat-glibc-2.1 packages have the system back up and running. and running fast.
The host is CentOS 7.3.1611 using bridged networking for KVM. So far it's running well.
On Apr 21, 2017, at 10:11 AM, Lamar Owen lowen@pari.edu wrote:
1.) Run Red Hat Linux 5.2 (or similar vintage) on KVM on CentOS 7;
For what it’s worth, I couldn’t get it working under a modern flavor of VMware, either. I find that telling because VMware tends to have the best driver support of all the VM systems, if only because it’s been around the longest.
Unfortunately, current VMware appears to have dropped Linux 2.0 support entirely, along with other contemporaneous things. For instance, even the “legacy Linux” version of its VMware Tools package contains Perl scripts that are written with the assumption that they’re running on at least Perl 5.8, which is contemporaneous with RHL 7.3 and kernel 2.4.18.
I spent some time trying to backport those scripts to the Perl 5.004 that ships with RHL 5.2, but gave up after making over a dozen changes with no obvious end in sight.
The way I see it, your solution involving CentOS 2.1 and the libc5 compatibility libraries just bought you the last upgrade to your software that you are likely to pull off without heroic efforts. I advise you to use CentOS 7’s remaining supported lifetime to get off this old software somehow.
You say there is no open source alternative, yet clearly the software was useful to at least you, and probably others, given that it appears to be commercial software. It might be prudent to sponsor the development of an open source replacement system.
e1000 drivers are not likely available (I couldn't find any).
That series of adapters didn’t get Linux support until about 2 years after RHL 5 shipped, according to the sources:
http://lxr.free-electrons.com/source/drivers/net/ethernet/intel/e1000/e1000_...
That would make the e1000 driver about ~3 years too late for you.
-
Lamar Owen -
Rainer Duffner -
Warren Young