Re: c++17,17,20... - was Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Nico Kadel-Garcia]

On Wed, May 5, 2021 at 6:22 PM Konstantin Olchanski  wrote:
>
> On Tue, May 04, 2021 at 11:00:00PM +0100, Andrew C Aitchison wrote:
> > On Tue, 4 May 2021, Konstantin Olchanski wrote:
> >
> > >- first slide of "distribution landscape" is nonsense,
> > >with everybody stuck with el7 for another 3 years and
> > >bye, bye, c++14, c++17, c++20.
> >
> > Is Red Hat Developer Toolset 10
> > https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_documentation_en-2Dus_red-5Fhat-5Fdeveloper-5Ftoolset_10_html_10.0-5Frelease-5Fnotes_dts10.0-5Frelease-23Changes-5Fin-5FDTS=DwIBAg=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Chv7ZrLnQII3dvv1o48KTg7YAxT9SNUUCkOkflNOCd4=bZUlWtCI03IJ95QWY-_DTkmqMdwOzXq8PVBkiR1dZLM=
> > not an option ?
> >
> > (OK, C++20 support in g++ 10.2.1 is "experimental).
> >
>
> And so what?
>
> I can take SL-6 and graft modern versions of all important packages,
> one does not even need the devtoolset, GCC is easy to build from sources.

Been there, done that, it's not as easy as you might think.
Backporting modern releases of Subversion or OpenSSH to RHEL 6 based
operating systems gets into dependency hell for critical libraries
fairly quickly, especially core system libraries like glibc and
gnutls. Don't get me *started* on the python dependency chains.

> But this is no longer "SL-6", it is "SL-6-KO1", at best.

It's common place with every major operating system. Sometimes the
base operating system lacks some critical components and dependencies.
I publish backports for Samba, and have done so for subversion, git ,
and openssh.

> Same thing, "CentOS-7 with devtoolset, php from webtatic, python from pip, 
> kernel from ELREPO, etc" is not CentOS-7.
>
> It is an irreproducible Franken-monster-bashed-together-locally thing.

And English has irregular verbs, and some people have some very
strange pronouns and peculiar colors of skin, or even wear glasses or
have embedded metal or plastic bits to replace joints or lenses or
even nerves.

> Is this the new standard, the best way to go, "the new thing" for production 
> environments?

I can testify with more than 30 years in the business that it always
has been done as needed. I used to staple components onto Red Hat
systems as needed for a network of roughly 13,000 hardened servers
over at Akamai roughly 20 years ago. It's not unusual.

Re: c++17,17,20... - was Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[~Stack~]

On 5/5/21 5:21 PM, Konstantin Olchanski wrote:

On Tue, May 04, 2021 at 11:00:00PM +0100, Andrew C Aitchison wrote:

On Tue, 4 May 2021, Konstantin Olchanski wrote:

[snip]

(OK, C++20 support in g++ 10.2.1 is "experimental).



And so what?

I can take SL-6 and graft modern versions of all important packages,
one does not even need the devtoolset, GCC is easy to build from sources.

But this is no longer "SL-6", it is "SL-6-KO1", at best.

Same thing, "CentOS-7 with devtoolset, php from webtatic, python from pip, kernel 
from ELREPO, etc" is not CentOS-7.

It is an irreproducible Franken-monster-bashed-together-locally thing.

Is this the new standard, the best way to go, "the new thing" for production 
environments?



I would say, no. The way forward is to use something like 
https://urldefense.proofpoint.com/v2/url?u=https-3A__spack.io_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Rcl3bOlhZYTsg6ao7N9s2R8gMaZj5RFHR3ZfE-XUUZg=XrKMgW2x7TS-6ye6hlykdflYSWiGTXaDqw3_WO5bTZw=  for reproducible builds of software. Or better yet, 
starting the difficult process of moving user applications into 
Singularity containers (https://urldefense.proofpoint.com/v2/url?u=https-3A__sylabs.io_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Rcl3bOlhZYTsg6ao7N9s2R8gMaZj5RFHR3ZfE-XUUZg=auYx06d5nuVcpRlexBCYemaHd0-4W213prqtSaLByHA= ). And getting Spack to build 
the Singularity images is even better! Both of those are fully Open 
Source tools with really good community support and free online training.


Once you can get the user applications into a container, you can 
abstract out the operating system (mostly; still needs to be Linux 
kernel - usually). Since Singularity is designed with HPC in mind, 
performance is fantastic.


We took an app that was built for RHEL 6, built it in a Singularity 
container, and can now run it on any Linux distro. As we move more of 
our user apps into Singularity containers we can start upgrading the OS 
and tools underneath the HPC environment without users ever knowing 
something changed (hopefully they notice the improvements).


Not saying that there isn't a learning curve for those creating the 
containers. I'm still not there in understanding it all and the 
container world is huge and varied. But it helped to just stick to 
Singularity and well establish formats until I got my head around it. 
And we haven't gotten to the point of letting users do it themselves 
yet. It's still an admin-only creation process. But we are getting there 
and the users don't have a clue how the app is installed/tweaked/tuned - 
they just know it works.


~Stack~

Re: c++17,17,20... - was Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

What you describe -- replacing a distro's utilities by those from other 
than the distro -- is done in practice for *SOME* things, as most on 
this list do.  However, under no condition should this be called a 
stable distro, let alone an "enterprise hardened stable" distro, without 
the amount of testing that most of us on this list realistically cannot 
provide.


I can provide examples.

Clearly, such "one-offs" are NOT production distros as you point out.

The real problem comes about when system core packages, such as "glib", 
Python, etc, in which the stock distro has a particular release level 
woven throughout much of the system core -- again, a "one-off" distro, 
but with all sorts of instabilities or even total failure.


Depending upon the access level any of the above has to other parts of a 
system, there could be large security vulnerabilities introduced.



On 5/5/21 3:21 PM, Konstantin Olchanski wrote:

On Tue, May 04, 2021 at 11:00:00PM +0100, Andrew C Aitchison wrote:

On Tue, 4 May 2021, Konstantin Olchanski wrote:


- first slide of "distribution landscape" is nonsense,
with everybody stuck with el7 for another 3 years and
bye, bye, c++14, c++17, c++20.


Is Red Hat Developer Toolset 10
https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_documentation_en-2Dus_red-5Fhat-5Fdeveloper-5Ftoolset_10_html_10.0-5Frelease-5Fnotes_dts10.0-5Frelease-23Changes-5Fin-5FDTS=DwIBAg=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Chv7ZrLnQII3dvv1o48KTg7YAxT9SNUUCkOkflNOCd4=bZUlWtCI03IJ95QWY-_DTkmqMdwOzXq8PVBkiR1dZLM=
not an option ?

(OK, C++20 support in g++ 10.2.1 is "experimental).



And so what?

I can take SL-6 and graft modern versions of all important packages,
one does not even need the devtoolset, GCC is easy to build from sources.

But this is no longer "SL-6", it is "SL-6-KO1", at best.

Same thing, "CentOS-7 with devtoolset, php from webtatic, python from pip, kernel 
from ELREPO, etc" is not CentOS-7.

It is an irreproducible Franken-monster-bashed-together-locally thing.

Is this the new standard, the best way to go, "the new thing" for production 
environments?

Re: c++17,17,20... - was Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Konstantin Olchanski]

On Tue, May 04, 2021 at 11:00:00PM +0100, Andrew C Aitchison wrote:
> On Tue, 4 May 2021, Konstantin Olchanski wrote:
> 
> >- first slide of "distribution landscape" is nonsense,
> >with everybody stuck with el7 for another 3 years and
> >bye, bye, c++14, c++17, c++20.
> 
> Is Red Hat Developer Toolset 10
> https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_documentation_en-2Dus_red-5Fhat-5Fdeveloper-5Ftoolset_10_html_10.0-5Frelease-5Fnotes_dts10.0-5Frelease-23Changes-5Fin-5FDTS=DwIBAg=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Chv7ZrLnQII3dvv1o48KTg7YAxT9SNUUCkOkflNOCd4=bZUlWtCI03IJ95QWY-_DTkmqMdwOzXq8PVBkiR1dZLM=
> not an option ?
> 
> (OK, C++20 support in g++ 10.2.1 is "experimental).
> 

And so what?

I can take SL-6 and graft modern versions of all important packages,
one does not even need the devtoolset, GCC is easy to build from sources.

But this is no longer "SL-6", it is "SL-6-KO1", at best.

Same thing, "CentOS-7 with devtoolset, php from webtatic, python from pip, 
kernel from ELREPO, etc" is not CentOS-7.

It is an irreproducible Franken-monster-bashed-together-locally thing.

Is this the new standard, the best way to go, "the new thing" for production 
environments?

-- 
Konstantin Olchanski
Data Acquisition Systems: The Bytes Must Flow!
Email: olchansk-at-triumf-dot-ca
Snail mail: 4004 Wesbrook Mall, TRIUMF, Vancouver, B.C., V6T 2A3, Canada

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Troy Dawson]

First off, IBM, the big company, has nothing to do with any of Red Hat
development.  We (Red Hat) are a completely separate company whose profits
go to IBM.  We have a completely separate legal staff, health care,
management, engineers, policies,  everything.  So, quit saying that we (Red
Hat) get access to all their lawyers and engineers.  We don't have access
to that any more than any other partner.  We also don't have to work
according to their policies, we have to work according to Red Hat's
policies, which are sometimes more strict, and sometimes less strict.

On Wed, May 5, 2021 at 8:32 AM Yasha Karant  wrote:

> You stated:
>
> all sorts of tests that we can't do in
> public, and make tweaks and changes that we can't do in public.  This is
> mainly due to hardware NDA's, and security stuff.
>
> End excerpt.
>
> 1.  Are these NDA enabled tweaks, changes, and tests that are in the
> production releases of IBM RH EL ("binary" installable "executables"
> only available under license for fee) fully and identically
> (defect-for-defect, etc) reproduced currently in CentOS (soon to be EOL)
> and thus in SL, etc., assuming that these are built from the IBM RH
> released source code?
>
>
All of Red Hat's source code is completely released, tweaks and all, when
the product is released.
Anybody can use that code.  CentOS, S.L., Alma, Rocky, Oracle, even you,
can freely download that code from the git repo's and do whatever you want
with it.

As was noted in a different email on this thread, you can freely get it
from the git repo's that are currently at git.centos.org.
Due to the EULA you cannot use the source rpm's that you get from your
subscription to build your own release.  That's a legal thing that makes my
head hurt.  But all the source that is those source rpm packages, is in the
publically available git repo's.  No subcription, or signup required.


> 2. "Security" -- you are including in this term the "hardening" of EL
> compared to, say, Fedora or "enthusiast" distributions.  If the
> "security" is done to any application that is derivative of a source
> code file that is under an "open systems" EULA, my understanding is that
> IBM RH would be required to release, not under NDA, the same security
> modifications.
>
>
All of Red Hat's source code is completely released, tweaks and all, when
the product is released.
See my previous answer.


> 2.1  Presumably, the large staff of legal professionals (eg, JDs)
> employed by IBM (and thus IBM RH) would do whatever is necessary through
> torts, actions, etc., to prevent the release of (1) or (2) to non-IBM
> entities if release is viewed to negatively impact the "financials" of
> IBM (including market share, cash flow, etc.).  Even if the "other side"
> has deep enough pockets to counter the above, by the time interval that
> the legal proceedings have concluded (unless these result in a permanent
> enforceable consent decree against the company and any of its successors
> -- tied to the intellectual property and the derivatives/successors
> thereof, not the current owner thereof), the modifications effectively
> are syntactically "obsolete".
>
>
2.1 is completely false, pretty much to every degree.

I already explained the IBM thing, but let me go back to my history, and
why I joined Red Hat.

I am one of the original creators and developers of Scientific Linux.
Connie and I were making Fermi Linux based off Red Hat Linux (RHL) before
Scientific Linux.
When Red Hat stopped RHL, and started Red Hat Enterprise Linux (RHEL), they
started a whole different process, with a whole different EULA.  No longer
would you be able to take their binaries and create your own distribution.
They were still working within the GPL, and the GPL gives them that right.

At the time, the other enterprise linux distribution was SUSE.  With SUSE,
they also followed the GPL, and if you were a customer, you could get
access to their source rpms.  They put alot of hurdles in the way, so even
as a customer, it was hard to get their source code.

We (Connie and I) were very worried about this happening at Red Hat.  But,
it didn't.

>From the very beginning of RHEL, all of their source for RHEL has been
publically available.  Originally it was via srpms at ftp.redhat.com, and
now in the git repos at git.centos.org.

Just think about that.   Here you have a new product, that you are asking
people to buy, and at the same time, you are giving away your source, in an
easy to consume way, in a public area.  It was very gutsy and courageous.
>From the very beginning, they knew what we (Scientific Linux, CentOS, White
Box ... etc) were going to do.  They were ok with it.

And rather than just withering away, this completely open policy, made them
grow, and continues to make them grow.

That was why I wanted to work at Red Hat.  They had courage, and they were
very open.

After joining, and working here for as long as I have, I still like it.
It's very open, and we continue to open things 

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Andrew C Aitchison]

On Tue, 4 May 2021, Yasha Karant wrote:

I fully concur -- a clear statement of a concern about the source and any 
authentication/pay-walls limiting access to that source.  I assume that the 
official rebuilders other than SL have paid the necessary fees to download 
the real, actual, production IBM RH EL buildable source as referenced below, 
currently at
https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_downloads_=DwIBAg=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=XlwZHatfU9lTVBkqWjIVraTktof8jaVDXPjoOzkoFGs=TX6gLDMRLMZ6f49TN5nbIEbiojZosZn3xgi4HyfiuR0= 
.
I do not have access -- but I assume that both Rocky and AlmaLinux have 
purchased the necessary access license.  I have not attempted to get a "dev" 
IBM RH license that supposedly is at no cost -- has anyone done so and down a 
full buildable source download?



From what I have seen on the EPEL mailing list, I think there are a few

bits missing. In addition to the code available with a devel licence
(which includes a "CodeReady Linux Builder" repo) in order to build 
CentOS the team had to get hold of a number of -devel packages on a 
case-by-case basis
https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.fedoraproject.org_archives_list_epel-2Ddevel-40lists.fedoraproject.org_message_P2DGC5KRDHT2JNGFGJEH3IZ24WTBIH4D_=DwIBAg=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=XlwZHatfU9lTVBkqWjIVraTktof8jaVDXPjoOzkoFGs=mPXFT8k9V2llxcjZMZ7XqyZXftuuZCcGUUY83jAntOs= 


--
Andrew C. Aitchison Kendal, UK
and...@aitchison.me.uk

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

You stated:

all sorts of tests that we can't do in
public, and make tweaks and changes that we can't do in public.  This is 
mainly due to hardware NDA's, and security stuff.


End excerpt.

1.  Are these NDA enabled tweaks, changes, and tests that are in the 
production releases of IBM RH EL ("binary" installable "executables" 
only available under license for fee) fully and identically 
(defect-for-defect, etc) reproduced currently in CentOS (soon to be EOL) 
and thus in SL, etc., assuming that these are built from the IBM RH 
released source code?


2. "Security" -- you are including in this term the "hardening" of EL 
compared to, say, Fedora or "enthusiast" distributions.  If the 
"security" is done to any application that is derivative of a source 
code file that is under an "open systems" EULA, my understanding is that 
IBM RH would be required to release, not under NDA, the same security 
modifications.


2.1  Presumably, the large staff of legal professionals (eg, JDs) 
employed by IBM (and thus IBM RH) would do whatever is necessary through 
torts, actions, etc., to prevent the release of (1) or (2) to non-IBM 
entities if release is viewed to negatively impact the "financials" of 
IBM (including market share, cash flow, etc.).  Even if the "other side" 
has deep enough pockets to counter the above, by the time interval that 
the legal proceedings have concluded (unless these result in a permanent 
enforceable consent decree against the company and any of its successors 
-- tied to the intellectual property and the derivatives/successors 
thereof, not the current owner thereof), the modifications effectively 
are syntactically "obsolete".


Yasha Karant

On 5/5/21 8:09 AM, Troy Dawson wrote:


On Tue, May 4, 2021 at 10:01 AM Yasha Karant > wrote:


If I correctly have read the RH EL9 CentOS announcement below, EL 9
will be in production before the end of this year, leapfrogging EL 8 as
it were.


I just want to clean up one point, because it seems you mis-understood 
the announcement.


RHEL9 will not be in "Production" before the end of the year.  It will 
be in "Open Development". [1]


Starting with RHEL8, RHEL now has 3 years between major releases. (8, 9, 
10), and six months between minor releases (8.4, 9.1).

RHEL9 will not go into "Production" until 3 years after RHEL8 was released.

But development of a release takes several years.
In previous releases, everything was done behind closed doors.  For 
RHEL8 I wasn't allowed to even say I was working on RHEL8.
For RHEL9, everything completely flipped open.  The goal was, and is, to 
have everything as completely open as possible.
If people want, I could go into each step, but let me make this brief 
and not do that.
By the end of the year, you, or anyone, can watch the whole process[2].  
Not just the "here is this days/weeks/months release".  But you will be 
able to watch, live, the RHEL developers git commits, merges, build, 
watch koji building the package, the package getting tagged, and so forth.
That's why I'm calling it "Open Development", because anyone will be 
able to watch the development.


So, what is the difference between that and a release?
We (Red Hat) do not stand behind the packages until they are released in 
production.  And only the packages released in production.
There will be a time that we take one of the snapshots, bring it behind 
closed doors and subject it to all sorts of tests that we can't do in 
public, and make tweaks and changes that we can't do in public.  This is 
mainly due to hardware NDA's, and security stuff.  That internal testing 
takes months.
After those tests and tweaking is done, and marketing says the time is 
right, we will release the production release of RHEL 9.0.


I hope this clears things up.

Troy Dawson

[1] - "Open Development" is my term, not a Red Hat term.  I find it more 
descriptive.
[2] - There are three exceptions to the completely open process.  1 - 
embargoed CVE's, 2 - code partners will not let us show until released, 
3 - some previously internal only package that got missed.  Number 3 is 
small, and hopefully should go away.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Troy Dawson]

On Tue, May 4, 2021 at 10:01 AM Yasha Karant  wrote:

> If I correctly have read the RH EL9 CentOS announcement below, EL 9
> will be in production before the end of this year, leapfrogging EL 8 as
> it were.
>

I just want to clean up one point, because it seems you mis-understood the
announcement.

RHEL9 will not be in "Production" before the end of the year.  It will be
in "Open Development". [1]

Starting with RHEL8, RHEL now has 3 years between major releases. (8, 9,
10), and six months between minor releases (8.4, 9.1).
RHEL9 will not go into "Production" until 3 years after RHEL8 was released.

But development of a release takes several years.
In previous releases, everything was done behind closed doors.  For RHEL8 I
wasn't allowed to even say I was working on RHEL8.
For RHEL9, everything completely flipped open.  The goal was, and is, to
have everything as completely open as possible.
If people want, I could go into each step, but let me make this brief and
not do that.
By the end of the year, you, or anyone, can watch the whole process[2].
Not just the "here is this days/weeks/months release".  But you will be
able to watch, live, the RHEL developers git commits, merges, build, watch
koji building the package, the package getting tagged, and so forth.
That's why I'm calling it "Open Development", because anyone will be able
to watch the development.

So, what is the difference between that and a release?
We (Red Hat) do not stand behind the packages until they are released in
production.  And only the packages released in production.
There will be a time that we take one of the snapshots, bring it behind
closed doors and subject it to all sorts of tests that we can't do in
public, and make tweaks and changes that we can't do in public.  This is
mainly due to hardware NDA's, and security stuff.  That internal testing
takes months.
After those tests and tweaking is done, and marketing says the time is
right, we will release the production release of RHEL 9.0.

I hope this clears things up.

Troy Dawson

[1] - "Open Development" is my term, not a Red Hat term.  I find it more
descriptive.
[2] - There are three exceptions to the completely open process.  1 -
embargoed CVE's, 2 - code partners will not let us show until released, 3 -
some previously internal only package that got missed.  Number 3 is small,
and hopefully should go away.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

If I understand your (Leon) interpretation of the IBM RH EULA, those who 
*START* from the actual, official, IBM RH source used for IBM RH EL, not 
that which is released currently as CentOS, violates that EULA.  Before 
CentOS became part of RH, later IBM RH, ("became part" is a WTO legal 
acquisition of intellectual property, such as logos, not necessarily 
personnel, etc), what did CentOS use?  If the only EULA allowed "source" 
for IBM RH EL is CentOS, not the actual IBM RH EL source, the only way 
to determine the fidelity of the CentOS source to the IBM RH EL source 
is for an entity (person?) who has a license to acquire the IBM RH EL 
source and do a string by string comparison.  The only differences 
should be logos and the like.  If the differences are build order, 
building methods, etc., one could envision latent defects and the like 
between the IBM RH EL installable "binaries" and the equivalent CentOS 
output file -- such latent defects should not be present if all 
compilers (including the "same" compiler save for release version) were 
defect-for-defect identical, and there were no loading/construction 
stream order defects (crudely, not revealed if A is built before B, but 
revealed if B is built before A).


Yasha Karant

On 5/5/21 3:17 AM, Leon Fauster wrote:

On 05.05.21 01:11, Mark Rousell wrote:

On 04/05/2021 23:41, Leon Fauster wrote:
The source are at 
https://urldefense.proofpoint.com/v2/url?u=https-3A__git.centos.org=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=0Cvrr_2WDkcsrPdHGtY_tjL0G9TG69QDuL7UWeyXUhc=ujZIUv5p_MGjr0gtRL9D7zIJAitkNauyt8wOXo1WxwQ= 



That's CentOS isn't it? That's fine if one wants to try and build 
CentOS but if one is building a RHEL clone (with Red Hat proprietary 
IP removed of course) then one would need the RHEL sources. The RHEL 
sources are available from 
https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_downloads_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=WwimKFFTyT3RShYiLFCGJpfGoLSUOFCE0z0_KBEyonY=TeQZDfODXKCZcEMTkVQuDIuZmOkbTDT3Q1tUmniPwKY=  
as long as you have a suitable account.


And, in any case, after CentOS 8 support ends then the only CentOS 
sources will be CentOS Stream sources which, as has been discussed at 
length, will not be suitable for many use cases.





Just to rephrase it: RHEL SOURCES are at the mentioned git server
and before RHEL7 it where on there FTP servers.

After EOL of C8 the sources should be still be pushed into git.

Using other SRPMS (like out of a RH Account) should be against
RH's EULA. Remember, when your intention is to build a clone.

Rocky is consuming via git.centos.org and Alma should also using it.
BTW, its questionable where the sources for AL8.4beta is coming
from ...?

AFAIK: The future will show if this all will be moved to gitlab ...

Thats so far my understanding.

--
Leon

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Lamar Owen]

On 5/4/21 3:38 PM, James M. Pulver wrote:

Maybe the AlmaLinux reddit (might also be one for Rocky?



Both Alma and Rocky have fairly active reddits.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Leon Fauster]

On 05.05.21 01:11, Mark Rousell wrote:

On 04/05/2021 23:41, Leon Fauster wrote:
The source are at 
https://urldefense.proofpoint.com/v2/url?u=https-3A__git.centos.org=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=0Cvrr_2WDkcsrPdHGtY_tjL0G9TG69QDuL7UWeyXUhc=ujZIUv5p_MGjr0gtRL9D7zIJAitkNauyt8wOXo1WxwQ= 



That's CentOS isn't it? That's fine if one wants to try and build CentOS 
but if one is building a RHEL clone (with Red Hat proprietary IP removed 
of course) then one would need the RHEL sources. The RHEL sources are 
available from https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_downloads_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=WwimKFFTyT3RShYiLFCGJpfGoLSUOFCE0z0_KBEyonY=TeQZDfODXKCZcEMTkVQuDIuZmOkbTDT3Q1tUmniPwKY=  as long as you have 
a suitable account.


And, in any case, after CentOS 8 support ends then the only CentOS 
sources will be CentOS Stream sources which, as has been discussed at 
length, will not be suitable for many use cases.





Just to rephrase it: RHEL SOURCES are at the mentioned git server
and before RHEL7 it where on there FTP servers.

After EOL of C8 the sources should be still be pushed into git.

Using other SRPMS (like out of a RH Account) should be against
RH's EULA. Remember, when your intention is to build a clone.

Rocky is consuming via git.centos.org and Alma should also using it.
BTW, its questionable where the sources for AL8.4beta is coming
from ...?

AFAIK: The future will show if this all will be moved to gitlab ...

Thats so far my understanding.

--
Leon

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Mark Rousell]

On 05/05/2021 00:38, Yasha Karant wrote:
> I have not attempted to get a "dev" IBM RH license that supposedly is
> at no cost -- has anyone done so and down a full buildable source
> download?

Just go here to create a free dev account (which allows up to 16 RHEL
instances and, of course, access to the source):
https://urldefense.proofpoint.com/v2/url?u=https-3A__developers.redhat.com_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=ObAgDY_TU7L4qolv1HO4YKSLeRAOSzDkm6OTpx47idw=8oCGvFp4dJb4JYifT2CdSkj6ZNstAyKIQHh0qmtXEoc=
 . I have a free dev account. It is very
easy to sign up.

Once you have an account you can easily download the sources. It is
obviously buildable if you can figure out how to build it. I've never
tried. Other clever people are kindly willing to do that for me!

The source (for RHEL 8.3, as an example) is downloadable as a 20.1GB ISO
containing, I believe, SRPMs. I did download it once but I've never
bothered to look inside it.

Building from the RHEL source RPMs (and removing the Red Hat proprietary
IP) is left as an exercise for the reader. ;-)

Thus building an open source clone of RHEL with the Red Hat proprietary
IP removed is of course possible, but no one said it had to be easy or
straightforward.

It is not surprising in my opinion that it has taken the Rocky project
some time to get to an initial RC state. As far as I am aware, Alma was
only able to do it more quickly because CloudLinux was already a rebuild
of RHEL (with their own customisations) and so they already had a
working build system in place that they could alter to produce a 'pure'
clone in the form of Alma.

> Does this also access the update buildable source for the installable
> "binary" updates provided by IBM RH EL for the licensees?

I think so, yes. There is an 'Errata' section in the downloads area for
RHEL that has a list of advisories. If you look at the details for each
advisory then you can see a list of updated packages for it and this
usually includes download links for both binary RPMs and source RPMs.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

I fully concur -- a clear statement of a concern about the source and 
any authentication/pay-walls limiting access to that source.  I assume 
that the official rebuilders other than SL have paid the necessary fees 
to download the real, actual, production IBM RH EL buildable source as 
referenced below, currently at

https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_downloads_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=j7mNaMr3RpyEdYeH8As2iHvK3COgKfyQyUEjt2ii5IY=p3k8zDAgC2P7lVIg-CwRQDgEYcVE4tshZ2NTxmm8ZDM=
  .

I do not have access -- but I assume that both Rocky and AlmaLinux have 
purchased the necessary access license.  I have not attempted to get a 
"dev" IBM RH license that supposedly is at no cost -- has anyone done so 
and down a full buildable source download?  Does this also access the 
update buildable source for the installable "binary" updates provided by 
IBM RH EL for the licensees?


On 5/4/21 4:11 PM, Mark Rousell wrote:

On 04/05/2021 23:41, Leon Fauster wrote:
The source are at 
https://urldefense.proofpoint.com/v2/url?u=https-3A__git.centos.org=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=0Cvrr_2WDkcsrPdHGtY_tjL0G9TG69QDuL7UWeyXUhc=ujZIUv5p_MGjr0gtRL9D7zIJAitkNauyt8wOXo1WxwQ= 



That's CentOS isn't it? That's fine if one wants to try and build CentOS 
but if one is building a RHEL clone (with Red Hat proprietary IP removed 
of course) then one would need the RHEL sources. The RHEL sources are 
available from https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_downloads_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=j7mNaMr3RpyEdYeH8As2iHvK3COgKfyQyUEjt2ii5IY=p3k8zDAgC2P7lVIg-CwRQDgEYcVE4tshZ2NTxmm8ZDM=  as long as you have 
a suitable account.


And, in any case, after CentOS 8 support ends then the only CentOS 
sources will be CentOS Stream sources which, as has been discussed at 
length, will not be suitable for many use cases.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Mark Rousell]

On 04/05/2021 23:41, Leon Fauster wrote:
> The source are at
> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.centos.org=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=0Cvrr_2WDkcsrPdHGtY_tjL0G9TG69QDuL7UWeyXUhc=ujZIUv5p_MGjr0gtRL9D7zIJAitkNauyt8wOXo1WxwQ=
>

That's CentOS isn't it? That's fine if one wants to try and build CentOS
but if one is building a RHEL clone (with Red Hat proprietary IP removed
of course) then one would need the RHEL sources. The RHEL sources are
available from 
https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_downloads_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=j7HyDpnhirEMJBg_NyqSf43v4NrMklLO4X7inAV3jK0=rd9KTZHBwuQQQ7W-Q27rrfW6d3PRUEOiLF2R0s3xmDg=
  as long as you have
a suitable account.

And, in any case, after CentOS 8 support ends then the only CentOS
sources will be CentOS Stream sources which, as has been discussed at
length, will not be suitable for many use cases.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Leon Fauster]

On 05.05.21 00:10, Mark Rousell wrote:

On 04/05/2021 21:42, Yasha Karant wrote:
Your statement at the end indicates that I have missed a source 
distribution channel.


Sorry, which statement is that?

Just for the avoidance of doubt, my comment about "Discourse" was a 
reference to the Discourse software > 
that Alma uses in place of a more traditional web forum.


You state that there will not be a CentOS 9 and only a CentOS Stream 
perpetual alpha or beta channel.


To be clear, I'm just an observer of all this but surely the whole point 
of Red Hat's infamous announcement is that CentOS 8 is to be the last 
version of non-Stream CentOS. So we know that there will not be a CentOS 
9. And we do know that there will be a CentOS Stream 9 because they have 
announced it.


I thought that IBM RH would not directly release buildable production 
EL source, but would channel it under a CentOS moniker.


To the best of my knowledge Red Hat never announced any change to the 
release of RHEL source in accordance with the software licences. If you 
have access to RHEL then you have access to the source ISOs.


How does IAS Springdale, Rocky, Alma, etc., get buildable production 
source for IBM RHEL9?


I would assume that (a) they download it from Red Hat using an account 
with legitimate access to RHEL (such as a free dev account), (b) modify 
it to remove Red Hat trademarked IP, and (c) design and operate a build 
system that allows them to build it. They will presumably get the RHEL9 
source ISO as and when RHEL9 is available.


I have a free Red Hat dev account (the same account type that has now 
been extended to cover 16 free RHEL licences) and I can freely download 
the ISOs containing the source.




The source are at https://urldefense.proofpoint.com/v2/url?u=https-3A__git.centos.org=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=0Cvrr_2WDkcsrPdHGtY_tjL0G9TG69QDuL7UWeyXUhc=ujZIUv5p_MGjr0gtRL9D7zIJAitkNauyt8wOXo1WxwQ= 




Does one have to buy the source from IBM RH? Will IBM RH or another 
IBM entity house the production source for the current production EL? 
What about the defect correction, including security defects, as well 
as minor release, update production source?


Nothing has changed to the best of my knowledge. It's all available 
according to the GPL terms. (Yes, I know that the RHEL software almost 
certainly contains code licensed under other software licences but GPL 
is certainly the main one that is of concern.)






--
Leon

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Mark Rousell]

On 04/05/2021 21:42, Yasha Karant wrote:
> Your statement at the end indicates that I have missed a source
> distribution channel.

Sorry, which statement is that?

Just for the avoidance of doubt, my comment about "Discourse" was a
reference to the Discourse software https://urldefense.proofpoint.com/v2/url?u=https-3A__www.discourse.org_=DwIDaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Rv6JPgfKB6WloyI9yzgIIJbr2llfrUuvqHQYxk1NNJw=h7okDDver9ejoWFtsZ0S8V58a9TPFpgQrea3hqgyB5U=
 >> that Alma uses in place of a more
traditional web forum.

> You state that there will not be a CentOS 9 and only a CentOS Stream
> perpetual alpha or beta channel.

To be clear, I'm just an observer of all this but surely the whole point
of Red Hat's infamous announcement is that CentOS 8 is to be the last
version of non-Stream CentOS. So we know that there will not be a CentOS
9. And we do know that there will be a CentOS Stream 9 because they have
announced it.

> I thought that IBM RH would not directly release buildable production
> EL source, but would channel it under a CentOS moniker.

To the best of my knowledge Red Hat never announced any change to the
release of RHEL source in accordance with the software licences. If you
have access to RHEL then you have access to the source ISOs.

> How does IAS Springdale, Rocky, Alma, etc., get buildable production
> source for IBM RHEL9?

I would assume that (a) they download it from Red Hat using an account
with legitimate access to RHEL (such as a free dev account), (b) modify
it to remove Red Hat trademarked IP, and (c) design and operate a build
system that allows them to build it. They will presumably get the RHEL9
source ISO as and when RHEL9 is available.

I have a free Red Hat dev account (the same account type that has now
been extended to cover 16 free RHEL licences) and I can freely download
the ISOs containing the source.

> Does one have to buy the source from IBM RH? Will IBM RH or another
> IBM entity house the production source for the current production EL? 
> What about the defect correction, including security defects, as well
> as minor release, update production source?

Nothing has changed to the best of my knowledge. It's all available
according to the GPL terms. (Yes, I know that the RHEL software almost
certainly contains code licensed under other software licences but GPL
is certainly the main one that is of concern.)

c++17,17,20... - was Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Andrew C Aitchison]

On Tue, 4 May 2021, Konstantin Olchanski wrote:


- first slide of "distribution landscape" is nonsense,
with everybody stuck with el7 for another 3 years and
bye, bye, c++14, c++17, c++20.


Is Red Hat Developer Toolset 10
https://urldefense.proofpoint.com/v2/url?u=https-3A__access.redhat.com_documentation_en-2Dus_red-5Fhat-5Fdeveloper-5Ftoolset_10_html_10.0-5Frelease-5Fnotes_dts10.0-5Frelease-23Changes-5Fin-5FDTS=DwIBAg=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=Chv7ZrLnQII3dvv1o48KTg7YAxT9SNUUCkOkflNOCd4=bZUlWtCI03IJ95QWY-_DTkmqMdwOzXq8PVBkiR1dZLM= 
not an option ?


(OK, C++20 support in g++ 10.2.1 is "experimental).

--
Andrew C. Aitchison Kendal, UK
and...@aitchison.me.uk

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Mark Stodola]

On 5/4/21 3:42 PM, Yasha Karant wrote:
You state that there will not be a CentOS 9 and only a CentOS Stream 
perpetual alpha or beta channel.  I thought that IBM RH would not 
directly release buildable production EL source, but would channel it 
under a CentOS moniker.  CentOS Stream is *NOT* production source.  
How does IAS Springdale, Rocky, Alma, etc., get buildable production 
source for IBM RHEL9?  Does one have to buy the source from IBM RH? 
Will IBM RH or another IBM entity house the production source for the 
current production EL?  What about the defect correction, including 
security defects, as well as minor release, update production source?



I encourage you to watch Troy's video posted earlier.

RHEL employees answer these questions.  CentOS Stream will be released 
as a new point release every 3 years, and end after 5 (same as RHEL 
general support).  CentOS Stream is the continuous integration that 
later becomes the minor point releases for RHEL. RedHat intends to 
continue to supply the RHEL source as they have been and is not making 
any effort to hinder the rebuilders (Alma, Princeton, Rocky, etc).

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

Your statement at the end indicates that I have missed a source 
distribution channel.  The discourse concerning the ergonomics of a list 
such as this versus an eyecandy "Ask Ubuntu" (with character count 
limitations on the length of postings) approach is worthwhile, but 
perhaps at a later time or off-list.


You state that there will not be a CentOS 9 and only a CentOS Stream 
perpetual alpha or beta channel.  I thought that IBM RH would not 
directly release buildable production EL source, but would channel it 
under a CentOS moniker.  CentOS Stream is *NOT* production source.  How 
does IAS Springdale, Rocky, Alma, etc., get buildable production source 
for IBM RHEL9?  Does one have to buy the source from IBM RH? Will IBM RH 
or another IBM entity house the production source for the current 
production EL?  What about the defect correction, including security 
defects, as well as minor release, update production source?


On 5/4/21 1:17 PM, Mark Rousell wrote:

On 04/05/2021 18:01, Yasha Karant wrote:
then one is forced to either Rocky or AlmaLinux, assuming either 
pushes out an EL 9 clone as soon as CentOS or other IBM RH buildable 
source is released.


Well, we know there's not going to be a CentOS 9. There will obviously 
be a CentOS Stream 9 but, as you say, that is not a viable replacement 
for those who want RHEL's stability without paying Red Hat for it.


The point of Rocky and Alma is to be downstream of RHEL so one assumes 
that they will do an EL9 clone as soon as they can after it exists.


Otherwise, for those who do not have a too heavy investment in 
hardware "driver" or specific software/systems application RPMs, there 
is Canonical Ubuntu LTS.  Ubuntu lacks anything similar to this list, 
as from my direct sign up and inspection of AlmaLinux does that distro 
as well -- both have something similar to "Ask Ubuntu" that is much 
more cumbersome and much more eyecandy than this straightforward 
list.  And, many more "non-systems" comments, much less of an 
"engineering" approach than this list.


The problem is that mail lists are so last century in the eyes of many. 
Personally I like mail lists but I know I'm in a diminishing minority.


There's nothing stopping a third party creating a mail list for Ubuntu 
or Alma but I regret to say that I doubt it would get many takers. Note 
that Ubuntu has a busy forum here, Ubuntu Forums 
, which is what people often desire nowadays. 
And Alma has its Reddit and a (dreadful in my opinion) Discourse group 
at AlmaLinux - AlmaLinux Discussion Community 
.


Some people love Discourse, others detest it. It seems very lightweight 
to me.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Mark Rousell]

On 04/05/2021 18:01, Yasha Karant wrote:
> then one is forced to either Rocky or AlmaLinux, assuming either
> pushes out an EL 9 clone as soon as CentOS or other IBM RH buildable
> source is released.

Well, we know there's not going to be a CentOS 9. There will obviously
be a CentOS Stream 9 but, as you say, that is not a viable replacement
for those who want RHEL's stability without paying Red Hat for it.

The point of Rocky and Alma is to be downstream of RHEL so one assumes
that they will do an EL9 clone as soon as they can after it exists.

> Otherwise, for those who do not have a too heavy investment in
> hardware "driver" or specific software/systems application RPMs, there
> is Canonical Ubuntu LTS.  Ubuntu lacks anything similar to this list,
> as from my direct sign up and inspection of AlmaLinux does that distro
> as well -- both have something similar to "Ask Ubuntu" that is much
> more cumbersome and much more eyecandy than this straightforward
> list.  And, many more "non-systems" comments, much less of an
> "engineering" approach than this list.

The problem is that mail lists are so last century in the eyes of many.
Personally I like mail lists but I know I'm in a diminishing minority.

There's nothing stopping a third party creating a mail list for Ubuntu
or Alma but I regret to say that I doubt it would get many takers. Note
that Ubuntu has a busy forum here, Ubuntu Forums
, which is what people often desire nowadays.
And Alma has its Reddit and a (dreadful in my opinion) Discourse group
at AlmaLinux - AlmaLinux Discussion Community
.

Some people love Discourse, others detest it. It seems very lightweight
to me.

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[James M. Pulver]

Maybe the AlmaLinux reddit (might also be one for Rocky?

--
James Pulver
CLASSE Computer Group
Cornell University




From: owner-scientific-linux-us...@listserv.fnal.gov 
 on behalf of Yasha Karant 

Sent: Tuesday, May 4, 2021 1:01 PM
To: scientific-linux-users
Subject: Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

If I correctly have read the IBH RH EL9 CentOS announcement below, EL 9
will be in production before the end of this year, leapfrogging EL 8 as
it were.  I wonder how much of this is due to the various issues with EL
8?  As SL 8 is not happening, SL 9 certainly is not -- forcing one to
choose to stay RPM or not.  If one stays RPM and does not want the
instability of CentOS stream (please see a previous posting to this list
with direct deployment observations of stream -- totally unsuitable for
a production hardened environment based upon what I read -- even less
"stable" than Fedora), then one is forced to either Rocky or AlmaLinux,
assuming either pushes out an EL 9 clone as soon as CentOS or other IBM
RH buildable source is released.  Otherwise, for those who do not have a
too heavy investment in hardware "driver" or specific software/systems
application RPMs, there is Canonical Ubuntu LTS.  Ubuntu lacks anything
similar to this list, as from my direct sign up and inspection of
AlmaLinux does that distro as well -- both have something similar to
"Ask Ubuntu" that is much more cumbersome and much more eyecandy than
this straightforward list.  And, many more "non-systems" comments, much
less of an "engineering" approach than this list.

On 5/4/21 9:46 AM, Leon Fauster wrote:
> On 04.05.21 17:41, Dave Dykstra wrote:
>> Yasha,
>>
>> I'll try to answer as I understand things as an observer.
>>
>> On Mon, May 03, 2021 at 02:51:30PM -0700, Yasha Karant wrote:
>> ...
>>> 1.  Fermilab and the non-CERN HEP community are not part of the Linux
>>> Future
>>> Community analysis as far as I can read, but are consulted later
>>> after the
>>> analysis is prepared (for HEP or CERN internal)?
>>
>> Fermilab and CERN have made it clear that they want to do everything
>> jointly.  They are considering input from the rest of the HEP community.
>>
>>> 2. CentOS Stream 8 repositories -- are these available outside of CERN?
>>> Outside of HEP?
>>
>> CentOS Stream 8 comes straight from Red Hat.
>
>
>
> JFI:
>
> "CentOS Stream 9 will launch in Q2 2021 as part of the RHEL 9
> development process."
>
> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.redhat.com_en_blog_faq-2Dcentos-2Dstream-2Dupdates=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=SfSn-uhVE6EuC2nmdoDVqqJNPWL1Ak06bk4NTVMleN0=
>
>
> Availability of Stream 9 packages on Gitlab, and a koji instance where
> you can watch package build activity.
>
> https://urldefense.proofpoint.com/v2/url?u=https-3A__blog.centos.org_2021_05_centos-2Dcommunity-2Dnewsletter-2Dmay-2D2021-2D2105_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=1_prbp5ptWOW4IlH1AtwMdgIqTByadUaIjE00obAC3A=
>
> https://urldefense.proofpoint.com/v2/url?u=https-3A__composes.stream.centos.org_test=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=dzekCNHUpo0DVAy9dFQWWJ38-TVG9bhnDuypLQJB2yU=
>
>
>
>>> 3. Note CC7 not SL7.  What are the differences?
>>
>> CERN and Fermilab did diverge on their approach to EL7.  You know what
>> SL7 is, and much of the HEP community stuck with that, but CERN based
>> their operating system on CentOS.  CC7 stands for CERN CentOS 7.  They
>> are basically compatible and I didn't hear of any application software
>> that noticed any difference beside the name.
>
>
>
> --
> Leon

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Mark Stodola]

Thanks Konstantin for all of this great information.

I'll second the recommendation to watch Troy's video.
It makes me feel a bit more comfortable that Alma/Rocky will be able to 
deliver the 10 years of updates.


I still have a sour taste for this new "module" packaging scheme. It 
seems to make rebuilding installation media with a custom package set a 
headache.  I've managed to hack something together, but it isn't 
pretty.  Has anyone else successfully tackled this?


-Mark

On 5/4/21 12:41 PM, Konstantin Olchanski wrote:

On Mon, May 03, 2021 at 09:14:11PM +, Dave Dykstra wrote:

Here's a presentation at HEPiX'21 from CERN that's publicly available:
 https://indico.cern.ch/event/995485/contributions/4256466/

Thank you for the link, I was not aware of this presentation.

For further reading, it contains a link to the March meeting of the linux
future committee: https://indico.cern.ch/event/1019875/

All materials are publically available, the zoom chat transcript requires
a CERN login (which I have), I am reading through the stuff now.


First impressions from the HEPIX presentation:

- there is no explanation how CentOS Stream is a technically acceptable 
replacement
for CentOS Linux. There is many questions here, none answered.

- there is no suggestion of reviving the CERN+FermiLab collaboration
(named SL or otherwise). This is surprising, if Princeton can roll out
a "new centos", CERN+Fermilab have even more resources to do same.

- there is no suggestion of CERN providing extended support for CentOS-8. To 
me, this would be the obvious path forward.

- first slide of "distribution landscape" is nonsense, with everybody stuck 
with el7 for another 3 years and bye, bye, c++14, c++17, c++20.

- second slide of "distribution landscape" has LHC experiments changing horses 
in the middle of Run 3. Good luck with that.

- no discussion of "red hat takedown/takeover of centos, 2.0" scenario for Rocky 
Linux & co.


First impressions from the "linux future" meeting in March:

- 30 minutes of our Troy Dawson leading Red Hat's video presentation: 
https://videos.cern.ch/record/2756480 (where they say "rail", they mean RHEL, 
took me a while to catch this)

- "CC7 is not an option to use for RUN3 at CMS", page 4, 
https://indico.cern.ch/event/1019875/attachments/2214410/3751654/lfc002.pdf

- support for 64-bit ARM is a requirement (good, but what about existing 32-bit 
ARM and 32-bit x86 machines?)

- page 5, key take away, "Most likely moving to CS8 is your best option, as this 
allows for a trivial upgrade path to EL8 at a later date".

(Is this true? Can somebody point me to the instructions for upgrading Centos 
Stream 8 to RHEL 8? A quick google search finds nothing)

- Linux BOF slides 
https://indico.cern.ch/event/995485/sessions/386343/attachments/2209929/3739897/hepixLinuxBoFMarch16th2021.pdf
- Linux BOF 82 min video https://videos.cern.ch/record/2756412

- Ben Maurice slides https://codimd.web.cern.ch/p/Zcc5CqncC#/1

- zoom chat (requires cern login) - all questions raised are reflected in Ben 
Maurice slides.


Bottom line.

All arrows point at CentOS Stream, but there is no review/evaluation of it's 
technical suitability
for HEP specific needs. Good/bad/etc compared to el7 and compared to non-RPM 
alternatives (ubuntu,
etc). And there is many questions, none answered.

It looks like everybody will have to do their own evaluation and decide for 
themselves.

P.S.

Do watch Troy's video!

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Konstantin Olchanski]

On Mon, May 03, 2021 at 09:14:11PM +, Dave Dykstra wrote:
> Here's a presentation at HEPiX'21 from CERN that's publicly available:
> https://indico.cern.ch/event/995485/contributions/4256466/

Thank you for the link, I was not aware of this presentation.

For further reading, it contains a link to the March meeting of the linux
future committee: https://indico.cern.ch/event/1019875/

All materials are publically available, the zoom chat transcript requires
a CERN login (which I have), I am reading through the stuff now.


First impressions from the HEPIX presentation:

- there is no explanation how CentOS Stream is a technically acceptable 
replacement
for CentOS Linux. There is many questions here, none answered.

- there is no suggestion of reviving the CERN+FermiLab collaboration
(named SL or otherwise). This is surprising, if Princeton can roll out
a "new centos", CERN+Fermilab have even more resources to do same.

- there is no suggestion of CERN providing extended support for CentOS-8. To 
me, this would be the obvious path forward.

- first slide of "distribution landscape" is nonsense, with everybody stuck 
with el7 for another 3 years and bye, bye, c++14, c++17, c++20.

- second slide of "distribution landscape" has LHC experiments changing horses 
in the middle of Run 3. Good luck with that.

- no discussion of "red hat takedown/takeover of centos, 2.0" scenario for 
Rocky Linux & co.


First impressions from the "linux future" meeting in March:

- 30 minutes of our Troy Dawson leading Red Hat's video presentation: 
https://videos.cern.ch/record/2756480 (where they say "rail", they mean RHEL, 
took me a while to catch this)

- "CC7 is not an option to use for RUN3 at CMS", page 4, 
https://indico.cern.ch/event/1019875/attachments/2214410/3751654/lfc002.pdf

- support for 64-bit ARM is a requirement (good, but what about existing 32-bit 
ARM and 32-bit x86 machines?)

- page 5, key take away, "Most likely moving to CS8 is your best option, as 
this allows for a trivial upgrade path to EL8 at a later date".

(Is this true? Can somebody point me to the instructions for upgrading Centos 
Stream 8 to RHEL 8? A quick google search finds nothing)

- Linux BOF slides 
https://indico.cern.ch/event/995485/sessions/386343/attachments/2209929/3739897/hepixLinuxBoFMarch16th2021.pdf
- Linux BOF 82 min video https://videos.cern.ch/record/2756412

- Ben Maurice slides https://codimd.web.cern.ch/p/Zcc5CqncC#/1

- zoom chat (requires cern login) - all questions raised are reflected in Ben 
Maurice slides.


Bottom line.

All arrows point at CentOS Stream, but there is no review/evaluation of it's 
technical suitability
for HEP specific needs. Good/bad/etc compared to el7 and compared to non-RPM 
alternatives (ubuntu,
etc). And there is many questions, none answered.

It looks like everybody will have to do their own evaluation and decide for 
themselves.

P.S.

Do watch Troy's video!


-- 
Konstantin Olchanski
Data Acquisition Systems: The Bytes Must Flow!
Email: olchansk-at-triumf-dot-ca
Snail mail: 4004 Wesbrook Mall, TRIUMF, Vancouver, B.C., V6T 2A3, Canada

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

If I correctly have read the IBH RH EL9 CentOS announcement below, EL 9 
will be in production before the end of this year, leapfrogging EL 8 as 
it were.  I wonder how much of this is due to the various issues with EL 
8?  As SL 8 is not happening, SL 9 certainly is not -- forcing one to 
choose to stay RPM or not.  If one stays RPM and does not want the 
instability of CentOS stream (please see a previous posting to this list 
with direct deployment observations of stream -- totally unsuitable for 
a production hardened environment based upon what I read -- even less 
"stable" than Fedora), then one is forced to either Rocky or AlmaLinux, 
assuming either pushes out an EL 9 clone as soon as CentOS or other IBM 
RH buildable source is released.  Otherwise, for those who do not have a 
too heavy investment in hardware "driver" or specific software/systems 
application RPMs, there is Canonical Ubuntu LTS.  Ubuntu lacks anything 
similar to this list, as from my direct sign up and inspection of 
AlmaLinux does that distro as well -- both have something similar to 
"Ask Ubuntu" that is much more cumbersome and much more eyecandy than 
this straightforward list.  And, many more "non-systems" comments, much 
less of an "engineering" approach than this list.


On 5/4/21 9:46 AM, Leon Fauster wrote:

On 04.05.21 17:41, Dave Dykstra wrote:

Yasha,

I'll try to answer as I understand things as an observer.

On Mon, May 03, 2021 at 02:51:30PM -0700, Yasha Karant wrote:
...
1.  Fermilab and the non-CERN HEP community are not part of the Linux 
Future
Community analysis as far as I can read, but are consulted later 
after the

analysis is prepared (for HEP or CERN internal)?


Fermilab and CERN have made it clear that they want to do everything
jointly.  They are considering input from the rest of the HEP community.


2. CentOS Stream 8 repositories -- are these available outside of CERN?
Outside of HEP?


CentOS Stream 8 comes straight from Red Hat.




JFI:

"CentOS Stream 9 will launch in Q2 2021 as part of the RHEL 9 
development process."


https://urldefense.proofpoint.com/v2/url?u=https-3A__www.redhat.com_en_blog_faq-2Dcentos-2Dstream-2Dupdates=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=SfSn-uhVE6EuC2nmdoDVqqJNPWL1Ak06bk4NTVMleN0= 



Availability of Stream 9 packages on Gitlab, and a koji instance where 
you can watch package build activity.


https://urldefense.proofpoint.com/v2/url?u=https-3A__blog.centos.org_2021_05_centos-2Dcommunity-2Dnewsletter-2Dmay-2D2021-2D2105_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=1_prbp5ptWOW4IlH1AtwMdgIqTByadUaIjE00obAC3A= 

https://urldefense.proofpoint.com/v2/url?u=https-3A__composes.stream.centos.org_test=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=dzekCNHUpo0DVAy9dFQWWJ38-TVG9bhnDuypLQJB2yU= 





3. Note CC7 not SL7.  What are the differences?


CERN and Fermilab did diverge on their approach to EL7.  You know what
SL7 is, and much of the HEP community stuck with that, but CERN based
their operating system on CentOS.  CC7 stands for CERN CentOS 7.  They
are basically compatible and I didn't hear of any application software
that noticed any difference beside the name.




--
Leon

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Leon Fauster]

On 04.05.21 17:41, Dave Dykstra wrote:

Yasha,

I'll try to answer as I understand things as an observer.

On Mon, May 03, 2021 at 02:51:30PM -0700, Yasha Karant wrote:
...

1.  Fermilab and the non-CERN HEP community are not part of the Linux Future
Community analysis as far as I can read, but are consulted later after the
analysis is prepared (for HEP or CERN internal)?


Fermilab and CERN have made it clear that they want to do everything
jointly.  They are considering input from the rest of the HEP community.


2. CentOS Stream 8 repositories -- are these available outside of CERN?
Outside of HEP?


CentOS Stream 8 comes straight from Red Hat.




JFI:

"CentOS Stream 9 will launch in Q2 2021 as part of the RHEL 9 
development process."


https://urldefense.proofpoint.com/v2/url?u=https-3A__www.redhat.com_en_blog_faq-2Dcentos-2Dstream-2Dupdates=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=SfSn-uhVE6EuC2nmdoDVqqJNPWL1Ak06bk4NTVMleN0= 



Availability of Stream 9 packages on Gitlab, and a koji instance where 
you can watch package build activity.


https://urldefense.proofpoint.com/v2/url?u=https-3A__blog.centos.org_2021_05_centos-2Dcommunity-2Dnewsletter-2Dmay-2D2021-2D2105_=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=1_prbp5ptWOW4IlH1AtwMdgIqTByadUaIjE00obAC3A= 

https://urldefense.proofpoint.com/v2/url?u=https-3A__composes.stream.centos.org_test=DwICaQ=gRgGjJ3BkIsb5y6s49QqsA=gd8BzeSQcySVxr0gDWSEbN-P-pgDXkdyCtaMqdCgPPdW1cyL5RIpaIYrCn8C5x2A=hYk1ITQcUtbqa-Vlm6RCnkOJbUbC15278L1IWzXyRqw=dzekCNHUpo0DVAy9dFQWWJ38-TVG9bhnDuypLQJB2yU= 





3. Note CC7 not SL7.  What are the differences?


CERN and Fermilab did diverge on their approach to EL7.  You know what
SL7 is, and much of the HEP community stuck with that, but CERN based
their operating system on CentOS.  CC7 stands for CERN CentOS 7.  They
are basically compatible and I didn't hear of any application software
that noticed any difference beside the name.




--
Leon

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Dave Dykstra]

Yasha,

I'll try to answer as I understand things as an observer.

On Mon, May 03, 2021 at 02:51:30PM -0700, Yasha Karant wrote:
...
> 1.  Fermilab and the non-CERN HEP community are not part of the Linux Future
> Community analysis as far as I can read, but are consulted later after the
> analysis is prepared (for HEP or CERN internal)?

Fermilab and CERN have made it clear that they want to do everything
jointly.  They are considering input from the rest of the HEP community.

> 2. CentOS Stream 8 repositories -- are these available outside of CERN?
> Outside of HEP?

CentOS Stream 8 comes straight from Red Hat.

> 3. Note CC7 not SL7.  What are the differences?

CERN and Fermilab did diverge on their approach to EL7.  You know what
SL7 is, and much of the HEP community stuck with that, but CERN based
their operating system on CentOS.  CC7 stands for CERN CentOS 7.  They
are basically compatible and I didn't hear of any application software
that noticed any difference beside the name.

Dave

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Yasha Karant]

Excerpts from the presentation referenced by Dave Dykstra are appended. 
There are multiple questions one may raise by reading the presentation; 
below are my initial questions.


1.  Fermilab and the non-CERN HEP community are not part of the Linux 
Future Community analysis as far as I can read, but are consulted later 
after the analysis is prepared (for HEP or CERN internal)?


2. CentOS Stream 8 repositories -- are these available outside of CERN? 
Outside of HEP?


3. Note CC7 not SL7.  What are the differences?

Start Excerpt.

CC7 (CERN CentOS 7) continues to be a supported operating system•
C8 (CentOS Linux 8) is available as a supported operating system

Users of CentOS SIGs (Special Interest Groups) will have no choice but 
to move to Stream (OpenStack, etc)


Factors affecting CERN for LHC Run313•
Both CC7 and CS8 distributions stop being supported mid Run3•
CERN IT Security will permit a small subset of firewalled hosts to 
utiliseCC7 or CS8 until the end of Run3•
Partially upgrading some resources to CS9 during the Run or during one 
of the “end of year stops” may be feasible for some workloads (e.g. WLCG)•
A decision on the successor of CC7/C8/CS8 needs to be made as soon as 
possible to allow experiments adequate time to plan for Run3


Linux Future Committee
Membership is internally comprised of representatives of key business 
units of CERN


Future Linux distributions will be RPM based•
Efforts required for retraining and retooling are too high to move away 
from RPM•

Future Linux distributions will need to support both x86_64 and aarch64•
Many (several thousand) SoC systems will soon come online with aarch64
[NB: aarch64 is 64 bit ARM, such as is used (I believe) in Fujitsu Fugaku]
CERN, Fermilab and other scientific sites are strongly interested in 
common roadmap(s)
Provide CentOS Stream 8 
repositories:•http://linuxsoft.cern.ch/cern/centos/{s8,s8-testing,s8-snapshots}

Long-term: 31.12.2021 onwards•
Drop support for CentOS Linux 8•
Support CentOS Stream 8 (until end-of-life 31.05.2024)•
Support “Next” Linux Operating system
Based on the analysis from the Linux Future committee, collaboration 
with Fermilab and other HEP communities, as well as the WLCG -decide on 
the path forward


End excerpts

Yasha Karant


On 5/3/21 2:14 PM, Dave Dykstra wrote:

Here's a presentation at HEPiX'21 from CERN that's publicly available:
 https://indico.cern.ch/event/995485/contributions/4256466/
My summary was much more succinct than the presentation and is partly my
own interpretation.  The presentation lists a whole bunch of options and
basically says that they're sticking with something related to RHEL,
will decide later which one, and in the meanwhile we can use CentOS 8
until the end of this year or CentOS 8 stream.

Dave

On Fri, Apr 30, 2021 at 05:03:20PM -0500, John R. Dennison wrote:

On Fri, Apr 30, 2021 at 09:35:02PM +, Dave Dykstra wrote:

Both Fermilab and CERN have stated that they plan to use CentOS 8 stream
for now (or Scientific Linux 7 or CentOS 7) and will evaluate later
whether or not to switch to one of the clones.


Can you please provide a citation for this?  The last I recall hearing
was what Konstantin referred to - that an official statement from
CERN would be forthcoming in Q1 of this year but I have yet to see such
a statement and am wondering if, perhaps, I overlooked it somehow.

Thank you,




John
--
"Be polite, be professional, but have a plan to kill everybody you meet."

-- General James "Mad Dog" Mattis

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[Dave Dykstra]

Here's a presentation at HEPiX'21 from CERN that's publicly available:
https://indico.cern.ch/event/995485/contributions/4256466/
My summary was much more succinct than the presentation and is partly my
own interpretation.  The presentation lists a whole bunch of options and
basically says that they're sticking with something related to RHEL,
will decide later which one, and in the meanwhile we can use CentOS 8
until the end of this year or CentOS 8 stream.

Dave

On Fri, Apr 30, 2021 at 05:03:20PM -0500, John R. Dennison wrote:
> On Fri, Apr 30, 2021 at 09:35:02PM +, Dave Dykstra wrote:
> > Both Fermilab and CERN have stated that they plan to use CentOS 8 stream
> > for now (or Scientific Linux 7 or CentOS 7) and will evaluate later
> > whether or not to switch to one of the clones.
> 
> Can you please provide a citation for this?  The last I recall hearing
> was what Konstantin referred to - that an official statement from
> CERN would be forthcoming in Q1 of this year but I have yet to see such
> a statement and am wondering if, perhaps, I overlooked it somehow.
> 
> Thank you,
> 
> 
> 
> 
>   John
> -- 
> "Be polite, be professional, but have a plan to kill everybody you meet."
> 
> -- General James "Mad Dog" Mattis

Re: [SL-Users] Re: any update on CERN Linux and CentOS-8 situation?

[John R. Dennison]

On Fri, Apr 30, 2021 at 09:35:02PM +, Dave Dykstra wrote:
> Both Fermilab and CERN have stated that they plan to use CentOS 8 stream
> for now (or Scientific Linux 7 or CentOS 7) and will evaluate later
> whether or not to switch to one of the clones.

Can you please provide a citation for this?  The last I recall hearing
was what Konstantin referred to - that an official statement from
CERN would be forthcoming in Q1 of this year but I have yet to see such
a statement and am wondering if, perhaps, I overlooked it somehow.

Thank you,




John
-- 
"Be polite, be professional, but have a plan to kill everybody you meet."

-- General James "Mad Dog" Mattis