Share on Twitter.
Get the most dependable SMTP service for your business. You wished you got it sooner!
June 4, 2014
Yet another critical security hole has been discovered and patched after researchers worked
out malicious servers could hijack Linux users of the cryptographic library.
Red Hat engineer Nikos Mavrogiannopoulos, who issued a security patch for the flaw (catalogued as CVE-2014-3466)
Saturday, shortly after it was reported May 28 by Codenomicon researcher Joonas Kuorilehto.
Users of other affected software will have to sit tight until their open source
developers incorporate the fix.
Until then, they'll remain open to potential malware attacks. The security vulnerability places GnuTLS back in the spotlight three months after Mavrogiannopoulos
discovered the dangerous bug affecting the Linux platform that allowed hackers to self-sign certificates
which would be gobbled up by targeted sites.
That critical security vulnerability affected more than 350 Linux software packages on
various distributions including Ubuntu, Debian and Red Hat/CentOS which depended on the GnuTLS
libraries for secure sockets layer and transport layer security.
The new security hole was described on the Red Hat bugtracker as a flaw in the way GnuTLS
parsed session IDs from Server Hello packets of the TLS/SSL handshake.
"A malicious server could use this security flaw to send an excessively long session ID value
and trigger a buffer overflow in a connecting TLS/SSL client using GnuTLS, causing it to crash
or worse, potentially execute arbitrary code," the description read.
"The security hole is in read_server_hello() / _gnutls_read_server_hello(), where session_id_len
is checked to not exceed incoming packet size, but not checked to ensure that it does not exceed
maximum session ID length.
A deeper analysis including proof of concept was published on the Radar.Today blog. Users
should update the latest GnuTLS versions (3.1.25, 3.2.15 or 3.3.4).
In other Linux and open source news
Red Hat wants to increase its market share in the storage software market, and has announced this morning that it has agreed
to acquire Inktank, a privately held developer of storage software.
The deal was done for $175 million in cash, and Red Hat expects to close the acquisition sometime in May.
Red Hat's fiscal 4th quarter earnings and outlook late last month proved a mixed bag, sending shares falling about seven percent.
This latest acquisition comes nearly three years after Red Hat snapped up storage software provider Gluster for $136 million.
Inktank is driving the adoption of Ceph, an open-source software-defined storage system designed to run on commodity hardware. Software-defined
storage and networks give software, rather than hardware, the leading role in network and storage management.
Red Hat's announcement of the Inktank acquisition included an assessment of the software-defined storage market from Ashish
Nadkami, an analyst for market tracker IDC says -- "Software-defined platforms will continue to grow faster than any other market
segment in the file and object-based storage market. This growth will primarily be driven by a rich and diverse set of data-intensive
use cases across multiple industries and geographies."
Overall, Intank will help Red Hat in expanding its storage capabilities, said Brian Stevens, Red Hat's chief technology officer.
"Inktank has done a very good job assembling a strong ecosystem around Ceph and we look forward to expanding on this success together,"
"The strength of these world-class open storage technologies will also offer compelling capabilities as enterprise customers
move to software-based storage systems," he said.
In other Linux and open source news
Since the recent discovery of the serious Heartbleed bug issue, members of the OpenBSD project have forked the popular OpenSSL
library with the goal of creating a new version that they say will be a lot more trustworthy.
And even though OpenSSL is still open source software, for a full two years its entire development community managed to overlook the
crucial bug that eventually triggered a global panic in the internet community.
The OpenSSL library has since been patched to address the security issue, but some fallouts from the crisis are still being felt, and
the single programmer whose error caused all the problems says there just aren't enough people scrutinizing the OpenSSL code to
clearly indentify difficult-to-find bugs such as Heartbleed.
However, the LibreSSL project wants to change all that, and fast. An actual fork of OpenSSL, LibreSSL was created by members of
the highly security-conscious OpenBSD operating system community, including its founder Theo de Raadt, who has publicly criticized
OpenSSL as a project "not developed by a responsible team." And we sure agree with him 100 percent.
The group's ultimate aim is to provide a drop-in replacement for OpenSSL that has been substantially rewritten and audited for
potential security vulnerabilities. The API won't change, they say, but much of the current code will, and in a rather drastic manner,
according to de Raadt.
But it's early days for the project yet, however, the group is moving rapidly. Its homepage says its contributors are currently "too busy deleting
and rewriting code to make a decent web page."
Much of the early work involves refactoring and cleaning up the OpenSSL code so that it's more readable and easier to maintain. A
quick glance at the code commits so far reveals a lot of "KNF" work – meaning the individual source files are being rewritten in "kernel
normal form," a standard C coding style used by BSD Unix operating systems.
Additionally, thousands of lines of unneeded and useless code have already been deleted as well. Much of that code was OS-specific,
including workarounds for such ancient platforms as VMS, OS/2, NetWare, classic Mac OS, and of course, older versions of Windows.
The good news of all this, however, is that LibreSSL will be an OpenBSD-only library. The developers do plan to provide multi-OS support
eventually, but only after they have rewritten enough of the code to make it stable and maintainable, and then find reliable developers
to work on ports to other systems.
"Now we all know that you all want this tomorrow," the project's homepage states. "We are working as fast as we can but our primary
focus is to deliver good software that we trust to run ourselves. We don't want to break your heart..."
As things stands now, the first version of LibreSSL is planned for inclusion in OpenBSD 5.6. If all goes according to plan, that iteration
should get here by sometime in November 2014. The upcoming version of the OS, OpenBSD 5.5, is due to ship on May 1st, 2014.
In other Linux and open source news
Linux and open source vendor Red Hat is hoping it will make a lot of money out of its OpenStack project near the end of next
year, and says it won't need the help of anybody to keep the project on track.
The company's remark as it refers to 'anybody' was meant as a jab to Linus Torvalds.
Red Hat said on Wednesday at its OpenStack summit that it will turn the data center management cloud technology into serious money toward the end of 2015.
Red Hat recently re-organized its business units to help it bring OpenStack to the next level in the enterprise segment, with
the hope of creating the same lucrative market for the data center management and provisioning technology as it did for Linux five years
"The next 1 1/2 year is paramount," explained Red Hat's general manager for OpenStack Radhesh Balakrishnan. "We already started ringing
the cash register on OpenStack. What do we see on the horizon? A ten to fifteen x scaling potential."
For the most part, the majority of Red Hat's OpenStack deployment has been for testing and development so far, Balakrishnan added,
but he expects major production, and with that major money, deployments to come along by the end of 2015.
Just like Linux itself, OpenStack will take several years to make money and Balakrishnan seemed to feel that the expectations by
the press for insta-profitability are a bit unrealistic.
OpenStack launched in mid-2010 with technology donated by the NASA and Rackspace, and since then has signed up a small list of contributors
including Intel, HP, Red Hat, and others.
"It's a three-plus year old startup," explained Dave Cahill, Solidfire's Director of Strategic Alliances when asked about what
he saw in OpenStack's future.
VMware, he pointed out, was founded pre-2000 and didn't start to make serious money till around 2009. A lot rides on OpenStack's
success as it gives company's a potentially cheap way of managing thousands upon thousands of servers without having to pay for the
"People want an easy-to-use tool to solve this and wish to get out of the VMware so-called tax," he explained. And OpenStack may
just be that tool. Time will tell.
One reason why OpenStack has failed to pull in as much cash as its various corporate backers hope could be a lack of focus within
the project that has led to feature-creep in some areas and a lack of development on key features like networking and scheduling
When we ask cloud insiders what could be done to give the project more focus, many argue that OpenStack needs a 'benevolent
dictator' who would lead OpenStack development in the same way Linus Torvalds uses his opinionated persona to steer Linux development.
Though this isn't a particularly pleasant way to develop software, having a single opinionated individual dictate the direction of
a project can give it focus. After all, besides Linux, many proprietary companies have grown successful by being led by a strong
leader like Jeff Bezos (Amazon), Bill Gates (Microsoft) or Larry Ellison (Oracle).
Red Hat isn't convinced by this argument and argues that OpenStack's "Foundation" model of governance is sufficient. "We don't
believe the need for a Linus figure," Balakrishnan said.
"Personally, if I had a choice between a reasonable set of customers and customer reasoning and someone who has a colorful personality, I'm
more convinced by the other. There is also the other dimension-- Linux was just compute, now you're talking about storage and networking and compute
and PaaS-- the scope gets really large for one single visionary," he added.
For OpenStack to be developed to the point of serious profitability "we need multitudes of leaders" Balakrishnan exclaimed.
For his part, Dave Cahill of Solidfire is optimistic as well, saying that the Foundation has "generally done a pretty good job" and
that OpenStack won't suffer as long as the Foundation "doesn't become a standards body."
"The OpenStack Foundation gives the technical meritocracy and influencing ability to the ones driving it. If the best brains are
driving it, then why worry about the personality?" argues Balakrishnan. And he might be right. We will see over time.
In other Linux and open source news
Canonical announced Wednesday that the latest long-term support release of its Ubuntu Linux distribution will be available Friday.
The South African company made the availability announcement of Ubuntu 14.04 LTS, codenamed Trusty Tahr yesterday, coincidentally
alongside chief rival Red Hat holding its Red Hat Summit in San Francisco. And no, that's not a coincidence.
The server features where Canonical is keen to gain influence against Red Hat is that Ubuntu 14.04 offers new functionality such
as support for the Icehouse OpenStack release, Containerization Technology Docker and ARM-64 chip support.
"We're seeing OpenStack more than anything," said Ubuntu server and cloud product manager Mark Baker. Canonical has also upgraded
its Metal-as-a-Service deployment technology to add support for more dense servers, such as AMD-SeaMicro's SM-1500 and equipment
from Cisco's UCS division.
There's also a series of incremental upgrades such as moving to support version 3 of the Puppet configuration management
software, upgrading the Xen hypervisor to 4.4, moving to support for version 0.79 of the Ceph object storage gear, and finally
supporting version 1.0 of LXC.
Non-x86 chip afficionados are also in for a surprise as well, with Ubuntu 14.04 supporting IBM's POWER 8 servers along with 64-bit
ARM chips from Applied Micro.
Database system admins may be relieved to hear that Ubuntu is offering some choice here, and is wrapping in some support for
MySQL 5.5 along with Maria DB 5.5, Percona XtraDB Cluster 5.5, and MySQL 5.6 into its distribution.
"Ubuntu is now the enterprise platform supported on the widest range of modern architectures-– IBM POWER, ARM64, x86, and x64," Canonical said
in its press release.
In other Linux and open source news
Red Hat said earlier this morning that it is lowering the price for accessing its publicly hosted OpenShift software as it
struggles to come to terms with the unstable economics of the cloud. The new Bronze pricing scheme for OpenShift was announced by Red Hat
in a blog post yesterday. It simply means that developers can now access an off-site version of the OpenShift PaaS (platform-as-a-service)
without having to pay a monthly fee, and instead only pay for the storage that they actually use.
"Overall, Bronze brings the real power of platform as a service by making it even easier to only pay for the extra resources you want
without a monthly platform fee," said Red Hat marketing VP John Poelstra.
Before the change, developers could either opt for a free version of OpenShift Online with limited amounts of infrastructure and
storage, or a $20 per month Silver option.
The Bronze package provides a halfway-point between the two, giving greater infrastructure than free, but lacking the Red Hat
support options of Silver, nevertheless.
"A segment of the users was either self sufficient or comfortable with the community based resources. We wanted to find a way to
provide some flexibility to developers who want to purchase and consume extra resources while utilizing community based support,"
said Red Hat's Director of OpenShift Online, Sathish Balakrishnan.
"Overall, a large percentage of the applications (1.6 million applications deployed to date) we have running on OpenShift Online are
found to have extra storage for both the application and/or database tiers.
Add-on storage gives these applications unlimited room to effectively scale and store data." OpenShift is a hosted platform-as-a-service,
competing with other remotely provisioned software like Cloud Foundry, CloudBees, Amazon Elastic Beanstalk, Engine Yard, and others.
The technology behind all of this is powered by Red Hat Enterprise Linux, and incorporates a few elements of SELinux and cgroups
to provide security and isolation for "gears", the containers in which all OpenShift apps run, and the fundamental unit of currency
for sizing an OpenShift installation.
Under the new pricing scheme, developers can access 16 gears of any size within the Bronze tier, and will need to pay $1 per
gigabyte per month for all storage they access above the 1 GB assigned to each gear.
In addition to this, users of Red Hat's Silver version, which costs from $20 a month, can now access more than 16 gears per OpenShift
Like Red Hat, OpenShift is built with a significant emphasis on open source and, because of this, competes with Pivotal's Cloud
Foundry project for the attention of Linux programmers and application developers.
Source: Red Hat.
Get the most reliable SMTP service for your business. You wished you got it sooner!
All logos, trade marks or service marks on this website are the property of their respective
companies or owners.
Linux News Today.org is read by over 450,000 people involved in the field of Linux application development,
professional Web hosting services, Linux
security, Linux Web development, etc.
Inquire about our reasonable advertising rates
on our news website. One of our advertising representatives will be in touch with you. Simply email us to learn
about our ad rates and how we can help drive relevant traffic to your website. Advertising space is limited.