Red Hat released RHEL 7 on 10 June. Highlights include:
- Docker (built-in automated software isolation through Linux Containers)
- the Linux 3.10 kernel
- XFS as the default file system
- secure access for Microsoft Active Directory
- the new init system: Systemd
- GNOME 3 is the default DE running in classic mode.
There is much more but be warned - the article has the details but it is RH's own press release so expect all the hype and spin.
This discussion has been archived.
No new comments can be posted.
Red Hat Enterprise Linux 7 Released
|
Log In/Create an Account
| Top
| 23 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(Score: 3, Interesting) by Subsentient on Wednesday June 11 2014, @02:20AM
Why not ext4? Why XFS? And as developer of Epoch [universe2.us], I obviously don't consider systemd an improvement.
"It is no measure of health to be well adjusted to a profoundly sick society." -Jiddu Krishnamurti
(Score: 2, Insightful) by Kunasou on Wednesday June 11 2014, @10:55AM
In Red Hat they love systemd, just look at Fedora.
(Score: 2) by tibman on Wednesday June 11 2014, @01:28PM
How many distros do you know of that have added Epoch to their package repository?
SN won't survive on lurkers alone. Write comments.
(Score: 2) by Subsentient on Wednesday June 11 2014, @04:22PM
One, my personal one. That's it. I'm hoping to improve those numbers. :^)
"It is no measure of health to be well adjusted to a profoundly sick society." -Jiddu Krishnamurti
(Score: 2) by Subsentient on Wednesday June 11 2014, @04:24PM
Ah, I misread that as "who is using Epoch as the default init system" to which my answer is, my personal distro. The answer to "who has Epoch available in their repos" would be Gentoo.
"It is no measure of health to be well adjusted to a profoundly sick society." -Jiddu Krishnamurti
(Score: 2) by tibman on Wednesday June 11 2014, @10:33PM
Why do you think more haven't picked it up?
SN won't survive on lurkers alone. Write comments.
(Score: 3, Informative) by ancientt on Wednesday June 11 2014, @07:17PM
XFS is faster for what RH expects their clients to be doing:
One of the big changes coming in RHEL 7 is the move from Ext4 to XFS as the default file system. Ext4 and its predecessor Ext3 have long been the default file systems in Red Hat's Linux distributions. However, Pacheco said, across all industries, Red Hat's customers are dealing with a data explosion - which shouldn't surprise anyone involved in the technology industry given the buzz around big data. "This enormous data growth requires a scalable, performance file system, which is provided by XFS - hence the move to it as the default file system," he said [internetnews.com].
It might help to know when Redhat recommends [redhat.com] choosing XFS:
Another way to characterize this is that the Ext4 filesystem variants tend to perform better on systems that have limited I/O capability. Ext3 and Ext4 perform better on limited bandwidth (< 200MB/s) and up to ~1,000 iops capability. For anything with higher capability, XFS tends to be faster. XFS also consumes about twice the CPU-per-metadata operation compared to Ext3 and Ext4, so if you have a CPU-bound workload with little concurrency, then the Ext3 or Ext4 variants will be faster. In general Ext3 or Ext4 is better if an application uses a single read/write thread and small files, while XFS shines when an application uses multiple read/write threads and bigger files.
This post brought to you by Database Barbie
(Score: 2) by aristarchus on Wednesday June 11 2014, @02:22AM
Hype and spin? Better than Fear, Uncertainty, and Doubt (now with Vaporware served in a cloud!), but don't see why this editorializing was called for.
(Score: 5, Insightful) by kaszz on Wednesday June 11 2014, @03:16AM
This site then missed these releases..
FreeBSD 10.0 [freebsd.org] released 2014-01-17
OpenBSD 5.5 [openbsd.org] released 2014-05-01 (the only BSD without secure homepage..)
NetBSD 6.1.4 [netbsd.org] released 2014-04-12
DragonFly BSD 3.8.0 [dragonflybsd.org] released 2014-06-04
And a lot of other software has been released. Does it change anything.. NO. So when you need software you go and check the homepage or repository and see what version is currently on offer. Should every software release be news. Well we would have no time to do anything else..
Transistors in the 100 GHz range that exists in labs. And would come closer to commercialization, that would change things!
Or software technique to make many operating systems 5x faster. Proven DoS proof socket service etc.
(Score: 3, Interesting) by aristarchus on Wednesday June 11 2014, @05:54AM
Make things better. Pass Turing test. Peta-flops achieved. Welcome to the singularity. (to be announced) But seriously, what would we do with all that processing speed? Solve Sodoku faster? But then what would we do with our leisure time.
(Score: 2) by kaszz on Wednesday June 11 2014, @11:03AM
I'm sure some medical and nuclear problem solving equations and simulations will have use for that processing power.
(Score: 2) by Geotti on Thursday June 12 2014, @01:06AM
What a silly question. What is the main driver of innovation in the computer world? Pr0n!
(Score: 0) by Anonymous Coward on Wednesday June 11 2014, @07:29AM
Given that OpenSSL seems to have more holes in it than a Swiss cheese, can we really consider any of the other BSDs' homepages "secure"?
Perhaps the OpenBSD team are just waiting until their fork of OpenSSL is ready for prime time -- after all "do the best with what we have" is a big part of the OpenBSD philosophy...
(Score: 2) by kaszz on Wednesday June 11 2014, @11:09AM
If they want to do the best with what they have. The project could add signatures embedded in the html-source. And TLS has been fixed by now? at least it will make life more difficult for TLA:s.
(Score: 2) by LaminatorX on Wednesday June 11 2014, @11:36AM
This site didn't exist for the FreeBSD release at the top there. ;)
While not every software release is newsworthy, RHEL is not just any software collection. It has a long term impact on many community members' work lives. For perspective, this story currently has the most views of the past ten articles. Clearly it was of interest to a lot of people.
I'm sorry you didn't find it compelling. Hopefully you'll like some of the other stories more.
(Score: 2) by kaszz on Wednesday June 11 2014, @01:15PM
That's a different perspective. Otoh the page views aren't public asfaik. Anyway that other dicey beta thing seems to present various software releases as news a little bit too often. So I hope that habit won't get entrenched here.
(Score: 2) by LaminatorX on Wednesday June 11 2014, @05:04PM
Yeah, if we ever have an entire topic category dedicated solely to Enlightenment updates, we'll certainly have gone too far.
(Score: 2) by kaszz on Thursday June 12 2014, @01:56AM
Does that dicey other site have that? ;)
(Score: 1) by digitalaudiorock on Wednesday June 11 2014, @02:08PM
I guess this would qualify:
I think I just threw up a little...not like I didn't know systemd was a forgone conclusion. What really irks me is that CentOS will obviously follow suit. All kidding aside, can anyone tell me anything that an enterprise server OS gains from Systemd?
(Score: 0) by Anonymous Coward on Wednesday June 11 2014, @02:51PM
I use Ubuntu, so I'm not sure how common those use-cases are (I don't use them - yet):
systemd has a fast dhcp built in (and some other infrastructure), so new docker instances (like a very lightweight, partially shared VM) can be brought up in less than half a second. You could add a new postgres instance with its own ip/port in about 2 seconds, for instance. A VM takes a few minutes to spin up.
(Score: 0) by Anonymous Coward on Thursday June 12 2014, @01:17AM
Repeat after me: "Containers are different from VMs."
(Score: 2) by caseih on Wednesday June 11 2014, @03:52PM
Absolutely. No more error-prone, fragile init scripts. No more hacked scripts to watch daemons and make sure they are functioning. Process supervision. I'd much rather write a 2 or 3 line system definition file than mess with init scripts running hundreds of lines that fail in weird ways when a dead PID file is found. But it's backwards compatible still, so if you want to mess with nasty init scripts you can. They still work.
Also, a much more detailed logging facility is here now, should you choose to use it (syslog is still there too).
There's a lot of FUD on sn about systemd, unfortunately, without really knowing much about it, and not wanting to know, which is a bit odd really.
(Score: 0) by Anonymous Coward on Wednesday June 11 2014, @05:51PM
it's beginning...
now our init system is written by the same misunderstood genius who brought us pulseaudio.
GREAT!
Time to switch to BSD until the Linux world gets tired of being fooled by the same guy.