Red Hat Linux 5.0
The biggest change in 5.0 is the switch to a different C library, Glibc. Linux 2.1.x kernel development is hopefully nearing the end of a long and exceptionally chaotic development cycle—the new dcache, finer grained SMP and a raft of new features and performance enhancements are causing trouble. This is nothing new for Linux, but it's been a long time since the last “stable” release of 2.1. On top of all the ongoing kernel changes is a new standard GNU C library, Glibc. Glibc has a clean threads implementation, transparent support for IPv4 and IPv6, improved linking and other incremental improvements that ultimately turn it into a whole new animal. The combination of new library plus new kernel code makes it harder to isolate bugs.
By biting the bullet now, choosing to bundle the new C library and stick with a proven release of the kernel, Red Hat is doing the world's Linux developers and ultimately its users a service. Threading is a standard feature of competing operating systems, such as NT and Solaris, and very important to Linux's new bleeding edge GUIs: KDE and Enlightenment. In the short term there will be a flurry of Red Hat 5.0 incompatible software which should be cleaned up by the time you read this article.
The graphical on-line help system is still hopelessly crude. The tools (notably htdig) exist to make an excellent searchable archive of all the included html/text documentation, and it's too bad that the default help system is so slow and clunky. The Red Hat 5.0 network configurator now has support for configuring a PAP (password authentication protocol) PPP session, which saves some troublesome scripting.
Red Hat 5.0 includes many security enhancements. Because of the change to Glibc, all of the applications had to be recompiled and relinked. As a result, it is hard to tell which programs have been upgraded or changed.
Yet, there are (as of this writing) 54 additional patches in http://www.redhat.com/support/docs/rhl/rh50-errata-general.html and 8 in Intel, including a critical kernel upgrade that blocks the Teardrop attack. The Teardrop Denial of Service attack immediately made the kernel (2.0.30) shipped with this release obsolete. Someone decided to hit me with Teardrop, crashing my servers every night for two weeks. The newest kernel (2.0.32) successfully detects, logs and blocks this attack.
You should immediately install a new kernel and many of the RPMs if you intend to use your 5.0 system on the Internet. On installing a new release, the first commands I run are:
mkdir /usr/local/rpms cd /usr/local/rpms ncftp -R -d 5 \ ftp.redhat.com:/pub/linux/redhat-5.0/ rpm -i -upgrade *
Substitute your favorite mirror for the FTP address. The wget utility (available from the GNU archives) is slightly better for mirroring sites in this fashion, as it supports both FTP and HTTP, partial file/directory transfers, time-stamped updates and automatically uses PASV mode (necessary for FTP to work through a firewall).
The fact that patches for these problems are so readily available is a tribute to the flexibility of RPM and the hard working bunch of Linux folk who are countering the persistent cracking community.
RPM's FTP and FTP proxy support are now documented on the man page, but to use RPM effectively to distribute your own software requires the book, Maximum RPMs, available from Red Hat.
The UNIX world needs a software installer of the caliber of the Windows 95/NT Installshield. RPM is a good, even great start, but Debian Linux's installer beats the RPM/GLINT combination for both interactivity and ease of use. An RPM/GLINT release combining the best features of RPM and Debian would be a good thing.
The RedBaron web browser has been dropped—Netscape 4.04 is freely available and is a vastly superior product. You will have to download it from the Net, however, as it's not bundled on the CD—yet.
Metro-X boasts of a nice configuration screen, added card support and improved color depth, resolution and performance over the XFree86 release of the X server. In my case, Metro-X is of marginal utility—the XFree and S.u.S.E. drivers for my ET6000 and Mystique cards are as fast or faster and more reliable. You have to configure XFree in order to even try to configure Metro Link in many cases.
Metro Link has some excellent add-on libraries, notably OpenGL, which is experiencing enormous interest in light of the PC gaming phenomenon. If you intend to use OpenGL, the bundled Metro-X server saves you money against the purchase of the extra OpenGL libraries. Your other choice is the freely available OpenGL clone, MESA, which converts OpenGL calls into X code—which is, in theory, much slower.
The new backup program with the unwieldy name of BRU 2000-PE is straightforward enough for a single-user environment. In my environment, where I have to back up multiple machines to a single tape drive, through a firewall, the combination of cpio/dump and ssh works exceptionally well.
The newly bundled RealAudio 5.0 server and client are very sexy products. I've now hooked up my studio to my main Linux box so that I can broadcast live sessions to the Internet. The server operated flawlessly. The live encoder worked great once I had a working sound card and worked out the correct encoding rate. The 95/NT versions of the software have nicer interfaces, but I've already had months and months of uptime on my primary RealAudio server.
|Speed Up Your Web Site with Varnish||Jun 19, 2013|
|Non-Linux FOSS: libnotify, OS X Style||Jun 18, 2013|
|Containers—Not Virtual Machines—Are the Future Cloud||Jun 17, 2013|
|Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer||Jun 12, 2013|
|Weechat, Irssi's Little Brother||Jun 11, 2013|
|One Tail Just Isn't Enough||Jun 07, 2013|
- Speed Up Your Web Site with Varnish
- Containers—Not Virtual Machines—Are the Future Cloud
- Linux Systems Administrator
- Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer
- Senior Perl Developer
- Technical Support Rep
- Non-Linux FOSS: libnotify, OS X Style
- UX Designer
- RSS Feeds
26 min 23 sec ago
- good point!
29 min 14 sec ago
- Varnish works!
38 min 21 sec ago
- Reply to comment | Linux Journal
1 hour 7 min ago
- Reply to comment | Linux Journal
3 hours 33 min ago
- Reply to comment | Linux Journal
7 hours 33 min ago
- Yeah, user namespaces are
8 hours 50 min ago
- Cari Uang
12 hours 21 min ago
- user namespaces
15 hours 14 min ago
15 hours 40 min ago
Free Webinar: Hadoop
How to Build an Optimal Hadoop Cluster to Store and Maintain Unlimited Amounts of Data Using Microservers
Realizing the promise of Apache® Hadoop® requires the effective deployment of compute, memory, storage and networking to achieve optimal results. With its flexibility and multitude of options, it is easy to over or under provision the server infrastructure, resulting in poor performance and high TCO. Join us for an in depth, technical discussion with industry experts from leading Hadoop and server companies who will provide insights into the key considerations for designing and deploying an optimal Hadoop cluster.
Some of key questions to be discussed are:
- What is the “typical” Hadoop cluster and what should be installed on the different machine types?
- Why should you consider the typical workload patterns when making your hardware decisions?
- Are all microservers created equal for Hadoop deployments?
- How do I plan for expansion if I require more compute, memory, storage or networking?