Recovery Time

Happy Returns

After a week at the seaside I returned home and immediately had computer problems. My trusty laptop who's hard drive I'd suspected being on the verge of failing, finally let go. I was in the process of trying to backup my data at the time when it repeatedly failed at the same file. So I ran a disk repair... diskutility reported a B Tree error, recommended I backed up my data, and then seemed unable to remount the disk. GREAT, how was I supposed to do that now?

I have a new disk on order, a Sandisk 480Gb Extreme SSD which has a read speed of 540MB/s and a write of 460Mb/s. That's nearly as fast as the mSATA card I installed in the new server. I'm looking forward to having faster access, longer battery life and hopefully better stability.

Not So Stellar

In the mean time I downloaded and installed Stellar Phoenix Mac Data Recovery. Initially it was just the free version, but when you see your apparently fine files listed you feel the impulse to grab your credit card and purchase a full license. So, I left it churning away for 24 hours, happily pulling hundreds of Gb of files off the damaged disk. Sadly when I tested them yesterday they all turned out to be garbage.

I've powered down the machine and restarted the attempt, but after 12 hours it's still reading the disk with not a file in sight. Yikes, this might be harder than I'd thought!

I'll just have to swap the disk out, stick it into a caddy and work on it later. I need the laptop working again, apart from everything else I use it to type blogs with!! I can't cope with just an ipad because I still use an ipod nano for audio books and podcasts, and I really need a proper keyboard and decent processing power for photo editing.

Going Live

The State of Things

My four wire case fan arrived last week featuring low speed operation, rubberised frame and unique corrigated design blades. Originally I'd ordered the three wire version by mistake, but this was unsuitable if you want the system board to be able control the fan speed. Then towards the end of the week I worked out how to transfer my databases over for Davical and Mediawiki. I'd already transferred my other mySQL databases by copying the files from /var/lib/mysql, but the Wiki failed to work and Davical uses PostgreSQL. I knew next to nothing about PostgreSQL so it was a slog to work out how the security worked, how data might be transfered, and how to fix the annoying missing PHP library problems along the way. Nothing works without a battle and this was a fresh install, so it should have been easy?! Go-live was now or never, the other apps could wait, and I didn't want to go through the hassle of transferring data again.

On the hardware side, the new fan (an 80mm 'be quiet!' SHADOW WINGS SW1)
required its mounting plugs to be modified, as the plastic locking pins wouldn't go through the threaded mounting bracket. It would have been flimsy anyway, so I discarded the pins, shortened two of the mounting plugs and used a couple of long bolts and washers to secure it. The fan was then plugged into the system fan header on the motherboard and the computer started. It was immediately obvious that the fan was running very slowly, the BIOS reporting just 500 rpm.

Fan mounted using modified only the lower pegs.

Nb, always test things before you modify them, that way if it's DOA you can get it replaced!

That should be just enough to keep a small amount of air moving through the case, and if temperatures heat up, then the fan should speed up to a maximum of 2000 rpm, at a whisper quiet 16.6dB sound level.

Out With The Old

Going live essentially means moving the data and backup disks over to the new server. But before doing this I was keen to compare power usage and disk speeds, (before and after) ultimately reassuring myself that spending all that money was worthwhile.

With all drives running on the old Via C7 based machine the system was shown to draw a steady 32 watts. This dropped to 24 watts after the drives were removed, leaving just the IDE SSD and case fan as additional loads. That's higher than expected for an eleven watt system board, even allowing for power supply inefficiencies.

I used a utility called hdparm to measure the drive speeds. This is a general purpose drive tweaking tool that can perform a multitude of tasks such as optimising speeds and changing time-out parameters. The following command produces two performance measurements:

   pingu:=# hdparm -Tt /dev/sdx (where x is the drive letter to be measured)

Drive	Cached Reads	Buffered Reads
IDE SSD	225Mb/sec	28.4Mb/sec
Samsung Momentus HN-M101MBB 2½" sata disk (via IDE to SATA adapter)	225Mb/sec	29.5Mb/sec
Samsung Momentus ST1000LM024 2½" sata disk (sata port)	233Mb/sec	30.1Mb/sec
Western Digital Caviar Green 3½" sata disk (sata port)	234Mb/sec	111.5Mb/sec

It's interesting that the SSD wasn't out-performing the other drives in this test, though in practice it had certainly yielded a performance boost when I'd first installed it. I know they wear out, but I'm sure they don't go slower!

For Shiela.

In With The New

The disks where installed in the new Intel i3 based machine and mount points added to the fstab. These I added using their Block ID rather than their device name, which is a unique drive identifier rather than one that's tied to the port being used. You can find out which disks are connected by typing the following command:

   pingu:=# blkid

The address numbers are a little bit too long to write down and type in manually, so I piped the output of this file into my fstab and edited the result to the correct format.

   pingu:=# blkid >> /etc/fstab

This task completed, I re-tested the disks for speed.

Drive	Cached Reads	Buffered Reads
Intel 525 60Gb mSATA card	6470Mb/sec	247Mb/sec
Samsung Momentus HN-M101MBB 2½" sata disk	6500Mb/sec	103Mb/sec
Samsung Momentus ST1000LM024 2½" sata disk	6590Mb/sec	100Mb/sec
Western Digital Caviar Green 3½" sata disk (in SATA2 port)	6500Mb/sec	119.5Mb/sec

There's quite obviously a performance boost but it's surprising that the 3½ inch desktop drive connected to a faster sata2 port wasn't any quicker on buffered reads. To be fair this disk is designed for efficiency and low power with only a 5,400 rpm platter.

Two laptop drives on top and a desktop drive under the mounting plate.

With all disks running the power drawn was just 23 watts, a nine watt saving over the previous machine. That's like turning off a compact fluorescent light bulb that's been burning for the last four years.

Power Corrupts

Pleased to Meter

My plug-in electricity usage meter arrived from Maplin (N67FU). It cost just just a tenner and gives me the ability to see if my careful purchasing has actually delivered me an efficient server. You plug it into a mains outlet, and then the device to be measured connects to the three pin 13 amp socket in the front. It has a maximum power rating of just over 3Kw so it should be enough for an Intel i3 (laffs), and the flat-ish design means wall-warts shouldn't be a problem.

It works really well, and it's cheap as chips, but the niggle I've had with it is reading the LCD screen. Parts of the display are really tiny and my eyes aren't what they used to be. Plus mains power sockets tend to be in low, poorly lit places, making reading harder still. The only real issue has been the mode indicator, which is the smallest bit and block inverse. (Who dreamed that up?.. I can understand from a design point of view how you'd want to differentiate that a little, but they might as have well used Egyptian Hieroglyphs!)

Meter mode indicator shown top right.

Thankfully the screen layout changes enough so that you can work out which mode you're in, and it's mostly about that big number in the top left so I can get by without needing a flashlight and my reading glasses.

Measuring Up

The existing Pingu had to be shut down to install a replacement /home and /data drive. I use 2.5" laptop drives these days to keep check on the noise and heat, and this disk was the first one I'd bought after switching to the policy four years ago. Last week it started to fail big time, throwing numerous errors, making loose rattling noises and finally showing files with zero bytes size. I won't bore you with the gory details because I have regular backups, but it gave me the chance to plug in the power meter and see how much juice my trusty old server has been using all these years.

The Via system board is rated at 11 watts, there's an SSD, two laptop drives and a 2Tb desktop drive for my auto-backups, which is normally asleep. So after allowing for a power supply efficiency of 80%, I'd estimated about 25 watts in total, but found it actually used just under 27w. I'm pretty sure it's the disks that are dragging it up so high, but until I unplug them it's just an educated guess. But having the electricity meter means I can be a little more pragmatic about future disk purchases. Maybe one day SSD's will be big enough and cheap enough to fulfil my data storage requirements.

Next to be benchmarked was the new Intel i3 based machine. The bulk of the software, two database servers (Postgresql and mySQL), as well as Apache2 had already been installed but not configured. And as I blogged last time, Minecraft Server was ready to go. Turning on the machine it peaked for a few seconds at 20 watts and then dropped to a very reasonable 14.5 watts. It pretty much stayed at that value until I started the Minecraft service and then it went up by about ½ a watt. I measured CPU temperature by installing LM-Sensors and CPU utilisation was monitored using the Top utility.

With one player joining the game the power rose to about 15.5 watts and (by pressing 1 while running top) I was able to see that all four of the CPU cores had started to register light utilisation. (So I stand corrected, Minecraft Server does use multiple cores - thanks to Dr Vesuvius for that heads-up).

top - 18:47:19 up 1 day, 22:48,  1 user,  load average: 0.18, 0.12, 0.08
Tasks: 106 total,   2 running, 104 sleeping,   0 stopped,   0 zombie
%Cpu0  :  5.6 us,  0.7 sy,  0.0 ni, 93.4 id,  0.3 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu1  : 18.6 us,  1.0 sy,  0.0 ni, 80.4 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu2  :  3.3 us,  0.0 sy,  0.0 ni, 96.7 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
%Cpu3  :  6.4 us,  0.3 sy,  0.0 ni, 93.3 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
KiB Mem:   8127440 total,  1199480 used,  6927960 free,    19984 buffers
KiB Swap:  3090420 total,        0 used,  3090420 free,   862192 cached

Some of this could be the other services running but without the Minecraft Server there's very little utilisation on the other cores.

With the case lid removed the CPU temperature drifted up to around 45 degrees centigrade and the fan still turned at it's original lazy rate. I gotta admit, that's better than I expected.

Coup de Grace

Digging in the Wrong Direction

Sometimes when you take a step into the unknown you fall into a hole. But when you start digging your way out, make sure you're not digging downwards! I'm starting to feel this way about the new server project. There's been new stuff to learn and as ever, the devil's in the detail.

I took the decision earlier this week to order a new mSata card, altering my return request from 'replacement' to 'money back'. The idea being to get alternative hardware much quicker, and once the returned item had been processed I'd be back in pocket in a week or so. Sure enough the order for the replacement card arrived the next day (another thumbs up for eBuyers free five day postage) and I called eBuyer's tech support to get a returns number, then printed off the postage label they'd emailed.

After seeing the newly delivered jiffy envelope my son was keen for us to try the server again that evening. We got the younger two off to bed, I set the server back up, whipped the lid off and fastened in the new card. To keep things simple I decided to just try a plain disk format rather than doing a partition copy. I ran fdisk, creating two partitions, and then entered the following command to format as the basic Ext2 linux format.

   pingu:=# mke2fs /dev/sda1

I'm using Ext2 because I want to limit writes to the SSD as writing to them wears them out. (No seriously it does!) Similarly I'll specify noatime in the fstab so that file last access times are not saved, further reducing disk writes.

The format command started OK, then I felt that awful deja vu as it crashed again with the same "Buffer I/O error on device" error, followed by a segmentation failure. There seemed little point in sending the original mSata card back now, clearly the problem lies elsewhere!

Search and Ye Shall Find

I did a bit of googling using search terms 'DQ77KB' and 'mSATA' and eventually I found Martijn Koster's blog about his almost identical build. The only real difference is that he's using a slightly faster processor. He'd put together quite a good overview but the thing that really interested me was that he'd updated his BIOS to 'KBQ7710H.86A.0051' using the F7 and USB stick method.

Well, in for a penny, in for a pound, I downloaded the zip version & did the same. Then I then restarted the computer and ran another format using the Debian 7 installer disk. This time it seemed to get through and went on to install the operating system. It still threw up the odd error along the way, but it allowed me to continue. After a few minutes it announced it was finished and then performed an auto re-boot. Sadly it but didn't get very far, more kernel errors flashed by and then it locked up. I was starting to wonder if something else might be wrong, maybe the memory?

I decided to sleep on it.  Z Z Z Z Z Z z z z z . . . . .

The following day I did a few more searches, mostly hitting the same stuff from the previous day, but then I saw a post where somebody recommended resetting the BIOS to defaults after after an update. Well that made sense, and I certainly hadn't, so I decided to give it a try that evening.

To my amazement it worked, so I returned back to the task of cloning my old disk and setting up lilo so the mSATA card would boot. Then remembering I'd previously altered some of the bios settings I set the RAM voltage back down to 1.35 volts. Restarting again I found that the computer would no longer boot, the errors coming back. This continued until I set the voltage back to the default  of 1.5 volts.

I'd gone out of my way to buy low voltage sodimms, but it looks like that may have been one of the root causes of my hardware problems. Now I had a 2nd nSATA card that I didn't need along with an mSATA to 1.8" adapter board (which I'd hoped to use as a backup plan). To add to my frustration the adapter came with a micros-sata connector on it (which was no use to me). Ghaahh.. how many connection standards do we bloody need!!!

1.8" Adapter board alongside standard SATA plugs

I looked up the micro-sata connection standard on wikipedia and found the data connection is the same but the power is much smaller. Instead to two separate plugs it expects one, with the data and power merged together.

Quick, Think of a Plan B!

The computer was now working fine, but cloning the old servers root partition wasn't. After messing about setting up the boot loader, all attempts to start it resulted in an invalid chipset error. I can't say that I'm that surprised, but it meant I was going to have to re-install Debian from scratch. It's not the end of the world, but it means running two servers until I got all the software installed and configured on the new box. I was really hoping to avoid this!

The first thing after a fresh operating system build was to install Java 7 and get Minecraft server running for 'His Nibbs'. Stack Overflow had a Q&A page where somebody had asked about how to install java using the apt-get utility. So I scrolled down to the answer, then copy and pasted the following code into the terminal window:

echo "deb http://ppa.launchpad.net/webupd8team/java/ubuntu precise main" | tee -a /etc/apt/sources.list
echo "deb-src http://ppa.launchpad.net/webupd8team/java/ubuntu precise main" | tee -a /etc/apt/sources.list
apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys EEA14886
apt-get update
apt-get install oracle-java7-installer

It worked a treat, adding the java sources and then performing the install in only a few minutes.

Next thing was to get the latest Minecraft Server build from www.minecraft.net/download. We grabbed the minecraft_server.1.6.1.jar and installed the linux screen utility. Requiring us to simply copy the jar file onto the server and run it with the following command:

screen java -Xms512M -Xmx1G -jar minecraft_server.1.6.1.jar

Then you can leave it running by pressing Ctrl + A D, and return to it later by entering screen -r

Is it me or do some sites seem to hide there stuff in amongst a load of old twaddle and psuedo-information? I scanned down the main minecraft.net page unsure of where to click,  then gave up and googled "minecraft server download" to find that link above. I despair at times!!

Designed to Confuse

Font of All Knowledge

It's a play on words, I don't mean "fount", and it's not about wisdom, rather the lack of. What I want to talk about is confusion within computer fonts... yeah I've found another problem that nobody else even considers to be one. Generally computers have brought beautiful typography to the masses. You try and draw letters on a poster without it, or use a typewriter to bash out a page of text, centering your titles and without making any mistakes. We have oodles of different font styles that we can use and (often) abuse, but strangely we tend to just stick to one or two.

The Arial font displayed in Windows Character Map.

Let's look for a moment at the characters in Arial, they look clean and familiar don't they. It's a hugely useful font designed in 1982 as a generic and bland version of a sans serif font. (Yeah somebody thought standard Sans Serif was too flashy!) Microsoft started shipping truetype versions of it in Windows 3.1 back in 1992, and it first appeared on the Mac when Apple released OSX in 2001.

Sans Serif means it doesn't have the small projecting features (called serifs) at the end of the strokes. Times New Roman is probably the most well known example with serifs.

Serifs shown here in red.

Undoubtedly the serifs looks grand, and they help differentiate some of the similar characters, but they have the disadvantage of being slightly less easy to read. They tend not to be used in large bodies of text, so fonts like Arial have become very popular, almost the de-facto standard because mostly people want clear readable text. But here's the rub, without the serifs some of the letters look the same.

 So What's the Problem?

It all came to a head recently when I was sent a new server password at work that had been created to fall in line with a rigid password policy. The policy states (something like) don't use real words, and include a mix of capitals numbers and characters. That's nothing unreasonable until you understand which characters had been selected to fit in with this policy.

The password was based on a meaningful term, in this instance system name and company abbreviation, as we outsource our IT hardware support to HP. The system was called TAMS and the company abbreviation is PLS, so the HP system administrator had unthoughtfully set the password to "T@msPl5"

OK, if you didn't know (& I didn't) what the password was based on then you might not be sure which letter was after the 'P'. "Big deal, just try again," you might say, but what if your policy is set to lock the account after three failed attempts? Non word passwords are difficult to type at the best of times.

I've had similar issues in the past when fellow developers had been slightly foolish in the selection of variables or labels in their code. One guy habitually used 'i' (for iteration) within his function loops, which was OK until he switched to using capital letters.

Still not 100% sure what I'm on about?.. well certain letters and numbers in fonts are either difficult or impossible to distinguish from one another. Look at the table showing comparison of some well used fonts.

Problem characters displayed using common fonts.

All of a sudden our Arial doesn't look so good. It tends to be mainly restricted to the fonts without serifs, and normally it isn't a problem because context helps us out. Where it becomes a problem is when you need to exactly identify the character when context is missing. Examples of this might be:

• Passwords - where a mixed character and uppercase policy is involved.
• Computer code or script - where single letter variable names are used.
• Serial Numbers - especially where the font size is tiny.

It seems that the problem is rooted in the past. Back in the days of moveable type you could reduce complexity by reusing the lower case L character for a 1. Indeed this practice was carried through to early typewriters which would typically exclude the number 1 from their keyboards. And fonts are created by designers, who care more about how something looks that whether it might cause the odd confusion between similar letters.

Avoiding The Problem

In years gone by when Car Registration plates in the UK used a single letter as a year suffix, certain letters where avoided. The letters I, O, Q, U and Z were never issued as year identifiers.

• I because of its similarity to the numeral 1
• O because of its identical appearance to a zero
• U because of similarity to the letter V
• Z because of similarity to the numeral 2

So back in years 1963 to 1983 they were aware of the problem and took proactive steps to reduce confusion and to assist with the correct identification on registration plates.So it's not rocket science, the boys at the ministry have been doing it for years.

Useful Methods

Not all solutions to problems have to be technical in nature, the common way we cope with such problems is to use comparison:

• Are there similar letters you can compare against, or similar examples that help identify the character you're unsure of?
• Are there any coding standards that prohibit certain characters?

Also you could consider the following:

• If you generate any codes, simply avoid any characters that might confuse?
• Start using a font like MS Trebuchet which retains character differentiation without resorting to serifs.

A Dim Outlook

I started looking into altering the font in my company email, but it only effects the outgoing emails that you create. Incoming emails are sent with the font pre-selected in what Microsoft calls the stationary, so it's not possible to change them. I found a tantalising stationary override tick-box in the setup page, but that disappointed me by doing nothing.

So I'm still stuck with copying and pasting before I can alter the font. Windows used to be all about control and configuration, what went wrong? I can't switch to another email client, it has to be outlook, yet it seems rigid in it's ability

Unless you know the answer!

A Game Of Two Halves

Building the Beast

It was the day of building the eponymous beast, we had all the bits, the house was cleared of small urchins with their touching fingers, and we were finally ready to go. I opened the system board box and carefully removed & inspected the contents. The Intel DQ77KB board comes with a poster sized "how-to-do" instruction sheet. It has good clear instructions, colour diagrams and each stage is split into simple numbered steps. My son Saul (who was new to computer building) had no problems following them under my watchful eye and with the odd word of assurance. I was like the sensei passing on noble skills to his apprentice.

Number one super guy!

We had a chat about static electricity and being careful about which bits you touch, and then kicked Off with Stage 1: processor installation. The Intel 1150 socket is a marvel of engineering and easier than it looked to use. I open and closed it a few times admiring the action, (it's a sort of tilt, slide and spring-lock). The retail boxed i3 processor came with a standard (four wire) cooling fan which just pressed into the four holes in the system board around the base of the socket. The fan header was close enough by to accommodate the short cable, and under operation it turned quite slowly and was almost silent.

nb. Inside the box I found a HDMI to DVI adaptor, which I'd gone out of my way to buy separately, but hey these things happen. The one I bought is far nicer :-)

Next the laptop style memory just slotted in like normal, though Saul was a little worried about how you had to joggle them a little before they would firmly snap down. The mSATA card fixes onto the full size PCI Express mini-card slot on the system board. I'd never seen these cards before and it really was small (see picture below with it in my son's hand). Resembling memory, it slid in at an angle and then had to be pushed flat against quite a springy force. Two tiny screws held it in place and there was some potential for this "liveliness" to catapult them (never to be seen again) if you didn't have your wits about you.

The mSATA card in Saul's hand.

Overall I am very impressed with the Intel system board, it's clear english instructions and the nifty processor socket and fan. Next step was to mount the board in the case and fire-it-up.

A Case of Beauty

The Streacom F7C EVO came exquisitely packed in a matt black cardboard box not unlike an Apple product, (we have a few of those in the house) protected by foam packaging and a white fabric bag. The comparison didn't end there, the look and quality of the case was equally top notch.

The Streacom F7C EVO Mini-ITX case.

I must say, I've very impressed by the solid feel and attention to detail. It's made of sand blasted thick anodised aluminium, laser engraved with large padded plastic feet and a minimalistic look. This does mean you don't get a reset button or a hard disk light. As a side project I started to contemplate adding an LED shining through the infra-red receiver window. The case is designed for a multimedia PC so an add-on receiver board for a remote control can be fitted behind the font panel, next to the two USB sockets.

Inside is also quite minimal (and no sharp edges anywhere, leaving fingers safe from cuts), there's a painted steel rack that bridges from front to back where you can bolt your hard disks (1 x 3.5" and 1 x 2.5") and a mount for a slot-loading CD/DVD drive. Then towards the rear there's a small bracket for mounting a standard 8" case fan.

Bagged Up Hardware.

Mounting hardware comes in individually labelled bags, but only sizes are stated & not what they're intended for. I took them out and by process of elimination worked out which ones where intended for mounting the system board. But I think they missed a trick here because there are no instructions in the box. Just adding 3.5" HDD or Motherboard to the appropriate label would have really helped, but if you've built a few machines before you'll soon work it out.

The motherboard matched the four mounting pillars exactly, but it does sit a little low in the case making some of the rear sockets slightly hard to access. Finally the power switch and power LED connectors were attached to the colour coded headers.

System Board Mounted Ready to Boot.

That went well, we had a full computer system on a board less than 9 inches square in under an hour.

And then the Fun Began

We both smiled as it booted first time, but then I couldn't get it into the BIOS settings. The splash-screen with instructions flicked by so quickly!! After a few restarts I figured out that F2 needs to be held down and then I was able to continue (surprising this wasn't on the quick-start sheet). I reduced the memory voltage down to 1.35 volts, and then spent a few moments looking at the various config pages. Each option has help text that displays over on the right of the screen, and it actually makes sense. This is a far cry from what I'm used to.

To take care of the disk cloning I downloaded a bootable CD image of Parted Magic. It took a while to start from a USB CD drive up on my old 1Ghz machine but once running it gives you a gnome based system to modify and manage your disks without having to get "down and dirty" with the command line. It actually does a whole lot more so I'd recommend checking it out if you run Linux.

Parted Magic Screen.

The plan was to clone the old IDE solid state disk onto a temporary disk, then clone that onto the mSATA drive. The first clone worked OK, taking about 10 minutes to do a 30Gb disk, but when I attempted the second stage (onto the mSATA card) it kept hanging. In one instance I left it running for over an hour, but it wasn't able to move on from 61% complete. I repeated this a few more times in vein, and then tried to manually format the mSATA from the command line. This rewarded me with a kernel panic and a "Buffer I/O error on device" message.

After a few more hours of trying different things I reverted back to my old server. I've raised an RMA with eBuyer to have it replaced, but it's left my son a little disappointed. He was really hoping to get the Minecraft server up and running. Perhaps it had been going too well, as those who have previously built computers will recognise, but as Saul's first experience at PC building it has actually been quite positive. He did tell me later that it was much easier than he'd imagined, and I think he enjoyed doing it.

So a result,.. though for the meanwhile only a minor one. (We'll try again after I've been through eBuyer's returns process)