There are two types of RAID, RAID that runs in hardware and RAID that runs purely from software. Hardware RAID is the more popular type, at least in terms of mindshare, with Hardware RAID controllers being nearly ubiquitous in PC platform servers. With Hardware RAID all RAID processing is handled by a dedicated RAID processor separate from the CPU with its own memory and often with its own cache. Software RAID handles RAID functions by using the system CPU and system memory to do the work and is generally either built into an operating system, part of a filesystem stack or available as a third party driver. Each type has its pros and cons (see the link for details) but far and away most people outside of the enterprise UNIX space with RISC platforms prefer hardware RAID for a variety of reliability and ease of use reasons.

What then is FakeRAID? FakeRAID is an odd animal that is truly RAID, but is software RAID pretending to be hardware RAID. In many cases it will be marketed as "chipset RAID" to make it sound like hardware RAID without really saying it. What makes FakeRAID really deceptive is that there is a hardware component of it to present the appearance of the RAID existing before an OS is loaded or installed (the chipset component.) Generally this means that the FakeRAID interface will be presented via the system BIOS or a separate BIOS at boot time. This makes it extremely difficult to tell if this is FakeRAID or Hardware RAID unless the hardware itself is investigated. To even a keen observer it looks exactly like normal Hardware RAID.

What FakeRAID does is use this BIOS functionality and interface purely to store settings in the BIOS that will be read by the Software RAID once it has loaded. But the Software RAID itself is just a driver that loads running on top of the operating system (and typically only Windows, there are no common examples of FakeRAID drivers working on non-Windows platforms.) So if something fails and the OS does not load or the driver fails, there is no RAID. And all RAID processing is done by the CPU and all RAID cache, if there is any, comes from system memory.

Since every enterprise operating system comes with software RAID built in (even Windows desktops commonly offer basic RAID options today) what you get with FakeRAID is truly nothing. It is an attempt to sell nothing, and they often get away with it as it is so confusing that even people who use it every day often cannot tell that they have been duped unless they are experienced with true Hardware RAID, are aware that a scam like this can exist and experience a system failure or measure system utilization to see that something is awry.

FakeRAID ends up being the worst possible scenario for RAID because it comes with all of the normal caveats of software RAID, adds in the obvious problems caused by confusion and misunderstanding by the system administrator(s) and has to do poor techniques in order to hide itself as a system driver which is the worst model for software RAID. With FakeRAID you lose the advantages that enterprise operating system software RAID can bring to the table and more. And, of course, because you are dealing with a product whose selling point is deceit or confusion you do not have a vendor with a focus on making a good product unlike the operating system vendors who have reputations to protect.

FakeRAID adds additional risk because of the bad vendor relationship it implies. Vendors selling FakeRAID have a relationship with you based on deception. If they are willing to do it with RAID, you can assume that they will do it elsewhere. FakeRAID means you and your vendor do not have a cooperative relationship, but they are in an active position of trying to trick you and take advantage of the situation. Your vendor is not operating ethically, nor with a production mindset, something absolutely critical for an IT production environment. Vendors selling you FakeRAID don't see you as a real business.

FakeRAID is sadly common and often hard to detect. FakeRAID vendors have a lot of incentive to hide the fact that their product does not do what it implies. In fact their whole business model rests on customers not figuring out that they are getting no functionality that they did not already have for free and that, worse yet, they are getting a worse version of what they already had for free!

We've updated to Ubuntu 23.10, the latest stable Node and the latest NodeBB platform. Check out all the new updates!

Just saw this and had to share. This is the most incredible picture of Spain. It's so good you can easily make out exactly where our village was where we lived earlier this year. This is probably the best picture of the peninsula that I have ever seen.

Doing a quick install of Ignite Realtime's OpenFire XMPP server onto a new CentOS 7 build (on Vultr) using MariaDB.

After a bare install of CentOS 7, updating all packages and logging in as root:

yum -y install http://www.igniterealtime.org/downloadServlet?filename=openfire/openfire-3.10.2-1.i386.rpm
yum -y install mariadb mariadb-server glibc.i686 
systemctl start mariadb
mysqladmin create openfire
cat /opt/openfire/resources/database/openfire_mysql.sql | mysql openfire;
firewall-cmd --permanent --zone=public --add-port=9090/tcp
firewall-cmd --permanent --zone=public --add-port=9091/tcp
firewall-cmd --reload
/etc/init.d/openfire start

Now you can browse to http://youripaddress:9090/ and complete the installation setup.

Today's news that the FreeNAS Coral project had to roll back their latest release due to instability is a great example of even one of the better and bigger storage "add on" projects having risk, and rather a lot of it, due to the Jurassic Park Effect. This is a truly massive setback for the project and will haunt them for a long time.

USPS: Unified SPam Service

@gbeyhaut said:

@JaredBusch prices on the web!! NEVER! www.cdw.com should have some prices. To be honest we sell these primarily through electrical distributors like Wesco and Rexel.

Even so, I think Jared is correct and getting MSRP prices or some guidance on pricing without needing to go to random online resellers. This presents a huge barrier to research and puts off IT departments. It works for IT that is being driven by their resellers, but not IT doing their own IT work.

What about a full day of Linux training in the training room? Like a class training, not a session.

@wirestyle22 said:

I tried to interview with CommVault and they started asking me specific CMD line commands I never use and host file editing, which I replied with "The only thing I really edit the host file for is to point to a different DNS and I've done that maybe one time in my life". They never called me back and basically stopped the interview there. I got pretty far with ABC like I said but that was more of a friendly conversation and ultimately they wanted someone with a lot of UNIX experience, which I basically have none. These experiences have kind of shaped my opinion of my knowledge.

So your takeaway here should be...

• Start learning PowerShell and if you get interviewed on CMD you throw your head back and laugh, mumble something about aging DOS era skills and do the thing that they want in PS.
• Start taking my UNIX admin course and do all of the stuff along with us as we do it. Get Linux running at home and get used to it.
• Keep interviewing and learning where you are weak and maybe get lucky when you find something.

Ansible, now owned by Red Hat, is one of the most popular, open source, DevOps management frameworks replacing cfEngine as the third of the big names in the space (along with the more well known Puppet and Chef.) As we use CentOS and Suse for most of our production servers and as Ansible is a Red Hat product, running it on Red Hat's CentOS 7 platform seemed to make the most sense. We are running this on our Scale hyperconverged cluster. Having a DevOps management toolset for a large scale virtualized environment is important for rapid building, changes, change tracking and more.

A default CentOS installation is going to have Python 2.7.5 which will work great. Ansible recommends getting the latest version from the EPEL repository but this does not work and Ansible has not updated the repo nor fixed the documentation. So we have to attempt installation through different means. We would use PIP, a Python package manager (similar to NPM for Node.) For PIP we will need the EPEL anyway, though. But PIP goes for an old version of Ansible too. Like the repos, PIP is on the Ansible 1.9 family when 2.0 and 2.0.1 releases are the current stables. So we need to go straight to GIT.

I like RPMs, they are easier to track, so we will use the GIT sources to build an RPM so that we get the best of both. Of course you could do this with a lot less effort skipping the RPM build, but since I am documenting it all and making it easy, there isn't any need to do so. This will allow RPM and YUM to report on Ansible being installed.

yum -y install git asciidoc rpm-build python2-devel
cd /usr/src
git clone git://github.com/ansible/ansible.git --recursive
cd ansible
yum install rpm-build/ansible-2.1.0-0.git201601101956.c3dd021.devel.el7.centos.noarch.rpm

That last line is for Ansible 2.1.0-0, which is the current version as of the time of this writing in the developer branch. You'll need to adjust this line to reflect the version that you have at the time that you do this.

That's it. Ansible 2 is installed. With the very latest developer release. Not ideal for you? Next we will modify this to get a stable release.

Jared Busch at MangoCon 2019

Youtube Video

Western Digital in Germany, always strangely out to promote RAID 5 because they use it in their entry level NAS products, tried to berate someone (again) for saying that RAID 10 was cheaper than RAID 5. WD conveniently leaves out things like risk and drive costs in these things and acts like parity is just cheaper than mirroring out of hand without actually looking into the issue. They are building quite a reputation for not knowing storage very well. This is the second or third ridiculously reckless posting from this in a week.

So I ran some numbers, using WD drives, to show why RAID 10 is cheaper. And it isn't a little cheaper. It is way, way cheaper.

RAID 10 Cheaper than RAID 5:

Current Amazon 3TB Drive prices:
3TB Red: £82
3TB RE: £202

To be even remotely safe in a tiny RAID 5, you need RE drives which are super safe and have very low UREs. Great drives. So three of those.

In RAID 10, there is no URE risk of array failure and so we can happily use Red drives at lower cost. But being RAID 10, we need four of them.

So the costs are:
RAID 5: £606
RAID 10: £328

RAID 10 looks to be significantly cheaper here.

Just saw @ryan-from-xbyte soft announce that xByte is going to be carrying HPE servers very shortly! That's very exciting, not that Dell servers aren't awesome but having the flexibility of both is a really big deal. He said that it is soon enough that quoting might be possible. Very cool.

PocketMine is a PHP-based server for the MineCraft Pocket Edition and MineCraft Windows 10 Edition clients. In this "how to" we are going to get PocketCraft set up to run as a service on CentOS 7.

I started this project with a minimal CentOS 7 template on a Scale HC3 cluster. In addition to the basic install, only firewalld and sysstat were added to the basic image. The image was fully updated as of the time of this writing.

For security reasons, PocketMine does not run as root and I think it would be best for it to run as its own, limited user rather than one used for any other purposes. Assuming that we stick with the default port of UDP 19132, we need to configure out CentOS firewall (firewalld and iptables) to accept traffic on this port. We start our configuration as the root user and will switch to our newly created pocket user once we have completed our root tasks.

useradd pocket
mkdir /opt/pocketmine
chown pocket:pocket /opt/pocketmine/
firewall-cmd --zone=public --add-port=19132/udp --permanent
firewall-cmd --zone=public --add-port=19132/tcp --permanent
firewall-cmd --reload
sudo -i su pocket

Now we have a user named “pocket” and we are acting as that user. This is much safer than using root or a standard user account.

cd /opt/pocketmine
curl -sL https://raw.githubusercontent.com/PocketMine/php-build-scripts/master/installer.sh | bash -s - -v development
./start.sh

By default the system starts with 256MB of RAM assigned to the PocketMine process. As this is a beefy server I decided to go ahead and give it 512MB as a starting point and monitor from there. The start script walks you through all of your setup parameters.

Now that we have a working server, we need to configure it to properly start on boot so that when the system reboots, the process will restart on its own without intervention. While still as the pocket user, edit the crontab with this command:

crontab -e

And enter this line:

@reboot	/opt/pocketmine/start.sh

And now you can reboot your server to test! If all is working correctly, PocketMine will be up and running and accepting requests once the server has restarts. Remember that if you are behind an external firewall that you will need to port forward UDP 19132 to the IP Address of your new PocketMine server so that clients can connect from over the Internet. If you only want to accept client connections from your local LAN you can skip this step.

Don't forget that if you want to publish this server to the Internet without needing to hand out a raw IP Address that you will want to make a DNS entry for it as well so that your users can connect conveniently.

PocketMine is a PHP application but is not part of a web server. So it uses the familiar PHP language platform but in a way that is foreign to most system admins, it is used as a traditional scripting language and is creating its own network services here.

You will notice in the docs that we use the --development flag for our server. This is obviously a questionable practice. Unfortunately because of changes to the protocol used on the clients that are out of our control the --stable branch does not connect, so at this time, development is necessary.

I use this anecdote from my own life quite often but do not have it anywhere to reference it, so I am adding it here as I feel that it is one of those really important stories that highlights "the cost of list opportunity" that can happen with university.

When I was about twenty I took a job working the overnight shift at a hotel. This was a great job because it was steady, paid a small premium (maybe one dollar per hour) over the normal desk clerk job because the shift was completely alone (ergo I was the overnight manager technically), it required doing the accounting and it was overnight. But in reality it was normally the quietest of the available shifts but it did require slightly more expertise than, say, the morning shift where you did no accounting and nealry always had someone there to assist with things.

I took the job purely as a way to pay the bills and loved that it left me free to do other things during the day, which I leveraged at first to be a struggling, but working, musician and later to start working IT short term jobs as they became available while paying the bills with something steady at night. It was not a career, it was just a job.

At the same time, I had many friends attending the local college (right across the street) who were studying hospitality management. Their goals, quite often, were to become hotel managers and felt that going to college was their path to that career. It takes, roughly, four years to get a degree in hospitality management if you attend full time and do not miss any classes. Those truly ambitious could speed that up and commonly it takes a little longer. But four years is common.

I did not do this alone, my roommate at the time was @AndyW and he worked at the same hotel that I did doing the opposite night's audit (we had some non-overnight shifts to fill in the gaps.) This is a handy way to work with a roommate because you often get twice the house because you work opposite shifts so overlap relatively little.

It took about eighteen months working at the hotel before both @AndyW and I were offered assistant manager positions. Just one and a half years. We both turned the offer down because the hotel was not a career for us but truly just a job until we made a break into IT (or music, we had a techno band!) We discussed it and knew that if we took the full manager position that we would start to feel locked in and making the jump to IT would involve too much pay cut and we likely would not do it. It worked out great for us, but my point is elsewhere.

The point here is that getting an entry point position in a hotel was easy. Hotels often struggle to fill these positions and turnover is high. From day one at the hotel we both had "overnight manager" to put on our resumes. Not very impressive, but not "desk clerk" either. In eighteen months, had we chosen, we were already being offered full time, day time, assistant manager jobs for the chain which is second in command to the general manager. My estimation is that without relocating I could have been in the position to be offered a general manager position in about two more years and if I had been willing to relocate anywhere for the position a bit less. That's all an assumption, but I was on the path to general manager very quickly. And had no college background to show for it.

My point here is that while this was happening, my friends who had their sights set on a career in hospitality were in college paying to be taught how to do the job I was already doing. Because they were done with their sophomore year, I was already at the full manager level. Chances are, I could have had my own hotel before they would have graduated.

And I did not start working in a hotel until I was twenty. Had hospitality management been a career goal I could easily have started doing some hotel work when I was sixteen, maybe younger and been a desk clerk the day I graduated high school and had even more jump on the people attending college. Coaching someone interested in hotels as a career from age fifteen, I could see someone being a general manager by twenty one with a solid background ready to propel them forward from there very quickly - a solid foundation.

The degree to which I had a head start over the college kids was amazing. I stumbled, entirely by accident, into their career goal before they had prepared themselves for their career start. And instead of paying money to a college and going into debt, I was making a respectable, if low, income during that time. I lived a comfortable life and enjoyed by job and used the free time that it gave me to train myself in another career field. I managed to build three careers (hospitality, musician and IT) all while skipping college and all of them past the point of the expectations of a college graduate's starting point - all while having misstepped as a late teen and getting a late start on my career path. Had I had any focus or had planned at all before being an adult I could have been much, much farther ahead still.

This is simply an anecdote, but an important one. So often college students, college grads or parents of students love to claim that there is no path into a field except through a college. And sometimes (doctors, lawyers, teachers in some states, pharmacists, etc.) this is true because of government regulations, but for most careers it is not true. It is repeated so much and so commonly accepted that people think it is acceptable to repeat. But examples exist over and over in many fields where alternatives don't just exist but are often very easy to execute - mostly because so few people bother to take advantage of them.

Hospitality management is a dramatic example where four year college grads are often looking to get the same entry level jobs as high school grads. The common belief is that only those with degrees get those jobs, but this is not true. Sure, those with degrees get top pick, normally, but those without degrees can, in theory, start picking many years earlier and build a career more quickly and be very far ahead.