Upgrading…

I remember back in the days where migrating to new software was anticipated by most.  People looked forward to getting the newer code that likely meant a more stable experience.  Stability was the primary reason for desiring the change, it seems.  They didn’t care that they would have to re-learn the changes incorporated into the software, as long as it meant that it would crash less and that they wouldn’t have to worry about losing data as a result.

It seems that since the days of Windows 2000 and Office XP, software reached a level of stability that most users became comfortable with.  Even to this day, Windows XP and Office 2003 seems to be the key choice among the corporate world.

Windows Vista and Office 2007 didn’t bring much to the table in terms of stability.  In fact, some could even argue that Vista is a step backwards in that area.  Office 2007’s main argument to use it over 2003 was that it’s new interface would increase productivity with the redesigned UI and Ribbon toolbar.  Of course, users would have to be re-trained, so it might be some time before that “productivity” starts to show.  :)

Then there’s the security argument.  A valid argument at that… for those who understand it.  Unfortunately, this ventures in the world of being something people don’t care about until it directly affects them in a largely negative way.  Kinda like your doctor telling you repeatedly to exercise, and you ignore it until you have a heart attack.

Forget the fact that Vista was poorly done.  Office 2007 wasn’t, yet the push to it was not nearly as urgent as 2003.  Why is that?  Software companies such as Microsoft are starting to become more aggressive about expiring support for older products.  One could argue their motives are financially driven.  Microsoft is just itching to pull the plug on XP, so they can begin to recuperate their losses from Vista.

Software has been coming to this point for quite some time now.  The kinds of things you use a computer for hasn’t drastically changed in the past decade or so.  If anything, they’re attempting to shift the focus into the cloud, but only to accomplish the same task.  Whether or not cloud apps will ever be able to overtake local apps is another argument.

It all comes down to what the end user wants.  It used to be stability.  Most people feel that software reached an acceptable point in stability on that timeline a while ago.  Their willingness to try something new just because of it’s prettier UI or couple of quirky new features they’d probably rarely ever user is just not enough anymore.  Security is not a good enough reason for them, and it never will be.  People will never be proactive about doing their own backups.  We’ve reached the point to where we, as IT professionals, have to start pushing end users to move forward, catching hell for it along the way.  Unless something changes in the world of IT, it’s only going to continue to be harder to do our jobs… not easier.  That sucks.

SSD Upgrade

I finally have a solid state disk in my unibody Macbook Pro.  I can, without a doubt, say this is the single best upgrade I have ever done on any personal computer… ever.  Forget RAM… get an SSD.  I used to have to wait ~30 seconds for my Mac to boot completely, login and all.  Granted, I have quite a few things that get started up at login.  Now, it’s maybe 10 seconds if that.  Apps launch instantaneously.  Decompressing large files is much quicker.  Everything is just faster… much faster.

Where I work, we are constantly dealing with large medical imaging datasets (such as MRI), which contain tens of thousands of smaller individual files.  So, as you can imagine, disk I/O is very important when cycling through these files for analysis.  At the rate SSD seems to be evolving, it won’t be long before our data processing workstations will all have one.  So, I needed one to test.  :)

I chose an OCZ Vertex 250G Mac Edition.  I did extensive research on which drive would be the best candidate to test, and OCZ seems to come out on top quite a bit.  Anandtech’s frequent reviews helped a lot.  The new Indilinx controller has been getting rave reviews, so my decision was easy.  If Intel had a 250G option, it would have made my decision much more difficult.

I’ve been using MacAuthority to source all of my Apple parts lately, and I’ve been extremely happy with their service.  I would even go as far to say it’s the best customer service I’ve ever received.  Next time you need something, give them a ring.  You won’t be disappointed.  So, I let them know what I wanted.  They didn’t carry that specific drive, but it seems all I had to do was let them know I was interested, and they got it for me.  I like that.  :)

So, the drive came in, and I wasted no time swapping it with my old 5400RPM disk.  I hooked the old drive up to a SATA to USB controller, booted from it, and proceeded to use Carbon Copy Cloner to image it onto the SSD.  Unfortunately, halfway through, the SSD disappeared and the clone failed.  ”Don’t do this to me… not with my new toy.”  So, I opened Disk Utility.. no SSD.  I rebooted, tried Disk Utility again… no SSD.  Tried ioreg -l -w 0 | grep OCZ, nothing.  Checked OCZ’s forums, and found another guy was having a similar problem and they recommended to remove the drive for 30 mins.  So I did, reinstalled it, it showed up!  The forum post recommended doing a firmware upgrade.  Downloaded the ISO, burned it, and rebooted.  It didn’t recognize the disk.  ”I can’t believe this”.  So, I contacted MacAuthority to let them know what happened.  ”OK, we’re sending you another one right away.”  Wow, am I a VIP or something?  :)  So, the new one comes in 2 days later.  I immediately check to see if it has the latest firmware, and it does.  I try the clone again, and it works fine this time.  Looking good!

Then I rebooted.  Wow.  I mean Wow.  I know I said it above, but wow.  I’m hardly ever wowed, but this time I was.  WOW.

I did before and after benchmarks using XBench.  If you search the Xbench submitted results for “SSD Upgrade“, you’ll see mine.  I titled both sumbissions the same, but I misidentified the Macbook Pro the second time so rest assured, they’re both me.  :)

Pretty definitive results:

• Pre-SSD – 37.02
• Post-SSD – 225.84

So, get one if you haven’t already (from MacAuthority!).  Sell a kidney if you have to.

I’m also looking forward to the next firmware upgrade, which is supposed to configure the drive to self-optimize with a garbage collection function and support for TRIM.  A beta version is on the forums as of this posting, but it seems that Mac owners are having problems with their Macs not sleeping properly, so I’ll wait.  :)

Social Media and Productivity

I’m a late bloomer in the whole social media scene.  I remember signing up for Twitter when it first came out, and thinking “I’ll never use this…”.  I still don’t use it very much at all.  I think I average ~5 tweets a day.  On facebook, I let my friends know what I’m doing maybe twice a month.  If you compare this to those who actually get paid for this stuff, there is NO comparison.  I’m a spec of dust in the twittersphere.  I still like doing it though.  I think it’s a great way to meet people, and branch out.

However, something alarmed me recently.  I do not get any monetary gain from posting to twitter, facebook, friendfeed, or my blog.  It is not a part of my job.  What alarmed me was, I’ve noticed I get a sense of accomplishment for doing these things.  I don’t think I should, as there is no real value in these things.  It’s a pasttime, a hobby, something to do when there’s nothing else to do.  What I’ve also noticed is, I have a tendency to try to squeeze these things in-between doing things I’m supposed to do.  This is bad when those things have a tendency to take longer than expected.  It’s all about your train of thought.  How easy is it for one small task to explode into an hours worth of reading?  It’s happened to me before many times.  I’ll click a bit.ly link on Twitter, which takes me to someone’s blog which I’m interested in, so I read the article, which also has several links that I invariably must read to fully encapsulate the purpose of the article, but then someone makes an interesting addendum or criticism in the comments along with their own links which must be read as well.  The next thing I know, I’m 50 years old.  Ok, that’s exaggerating, but you get the point (time flies).  I’ve easily wasted entire days engrossed in absorbing information, that I’ll look at the clock “5 minutes from when I sat down that morning AM” only to see hours have passed.

However, I feel like I accomplished something by doing that.  I learned something I didn’t previously know.  The question is, will I ever actually use that knowledge?  Does knowledge always need to be 100% useful to make it worthwhile?

I try to keep my information inflow trimmed to the bare minimum, which is already too much as it is.  Is this an acceptable way to spend time?  I hear a lot of “you just have to manage your time”, but I think that falls more along the line of “you just have to eat healthy”.  Everybody knows they need to do it, but very few actually do.

I’m not sure that makes sense, or if I accomplished everything I originally meant to say.  Too many distractions.  Gotta go.

RAID Mirror on OS X Server

We have two production Tiger servers, and I decided I wanted to get RAID1 set up on the Server HD itself.  Apple tells you this isn’t possible without reformatting/reinstalling.  AFP548 tells you otherwise:  http://www.afp548.com/article.php?story=20040827122302975

However, I spent several hours trying to get this operational with no success.  If you look in the comments of that article, I’m not the only one.  I was probably leaving out an important step somewhere, but after a few tries of reformatting, etc, you get a little brain fuzz.  I’m sure being there at 2am wasn’t helping either.  :)

I decided against doing this for now.  Instead I opted for a nightly incremental clone using Carbon Copy Cloner to the second HD.  In hindsight, this is probably going to serve me better.  Let’s say an update goes awry and renders something unusable.  I still have yesterday’s clone to boot to!  Hardly anything changes on the server HD itself, since these servers are just file servers, so that works out well.  I can see this not being a good setup for some depending on what kind of server you’re running and where the important data is actually stored.  In my case, we have 3 Apple RAIDs at ~17TB total (older, but still running), and just bought an Active XRAID 16TB to compliment that 17TB we already had.  All configured in at least RAID5, with the Active XRAID set up with 2 dual parity RAID6 slices.  Plenty of protection there.  And of course, RAID is NOT backup!!

When our next server comes in, I’ll probably set up a RAID1 mirror with a CCC clone to a single 3rd drive, or maybe use the 3rd drive as a hot spare.  I’m liking the idea of having an operational 1 day old backup, so the first is more appealing.

On a side note, I’m interested to see what RAID storage offerings Apple will be offering now that Snow Leopard Server will do ZFS*.

*Apparently, Apple seems to have removed ZFS from it’s Snow Leopard page.  Say it isn’t so!!

OpenSolaris 2009.6 in VMware Fusion

To install vmware-tools:

ln -s /usr/lib/vmware-tools/configurator/XOrg/7.1 /usr/lib/vmware-tools/configurator/XOrg/7.4

or if you’re using 64-bit:

ln -s /usr/lib/vmware-tools/configurator/XOrg/7.1_x64 /usr/lib/vmware-tools/configurator/XOrg/7.4

Taken from here:  http://communities.vmware.com/thread/212833

**Edit**

This fixes the installation problem, but the tools themselves still do not work properly.  I haven’t found any other solution yet.  I’m going to try open-vm-tools next.

Search

Lots of news on the Search front lately.  Wolfram Alpha released their computational knowledge engine.  Microsoft announced their upcoming Bing search engine which is supposed to replace Live search.  Do any of these pose a threat to Google?  I think no.

I don’t think Wolfram Alpha ever intended to challenge Google.  I won’t use that search engine for anything other than if I need a scientific explanation of something, or maybe a graph plot of some mathematical equation.  If anything, I can see Wolfram Alpha being a huge compliment to Google.  Do I smell a buyout?  That would actually be pretty nice!  Considering Google’s stance on education, a Wolfram partnership would go a long way.  Unlike Microsoft, any company or product Google buys actually flourishes.

Bing is trying to be more of a direct threat to Google.  It’s no secret that Google is one of Microsoft’s biggest competitors.  The reason being is typical of Microsoft history.  They tried to take someone else’s idea and make their own version, and assumed people would prefer that version just because it’s made by Microsoft.  In this case, they took too long to come out with something comparable to what Google had.  That’s why Google is synonymous with the term “search” now… “Just Google it”.  What do you think would happen several months from now if somebody said “Just Bing it”?  I would laugh.  Google is already the search king, and has been for over a decade.  People associate the term “search” with “Google”, and that is why it will stay on top.

Since then, Microsoft has been playing catch-up with Google, copying idea after idea… trying to offer a comparable product and some competition.  I personally don’t know anyone who uses Microsoft Live anything.  I’ve signed up for most of their stuff, but only to check it out.  I never use it.  It’s just not cross-brand compatible.  True to Microsoft’s corporate identity, they make it easy to use their stuff with their own products… not so much with anything else.  Google is pretty much brand independent.  They have the luxury of being so because everything they offer is the same.

I have no idea if Bing is actually better than Google.  From the reviews I’ve been reading, it seems like it is.  <conspiracy_theory_alert>That is, of course, assuming the people reviewing haven’t been paid off by Microsoft!</conspiracy_theory_alert>.  This isn’t going to matter, for three reasons:

1. Bing is still made by Microsoft and for that reason alone, most people will not like it due to Microsoft’s corporate identity.
2. Google’s search engine is due for a major update.  They announced Wave, but that’s not it.  I have a hunch something else is coming.  Something big.
3. Google=Search.  That will be very hard to overthrow.  For the same reason that for the longest time, Microsoft=PC.  Also, I don’t think Google will make the same mistakes Microsoft did.

The only thing Microsoft has going for it is it’s market share.  While still at a staggeringly high 87.90% as of this posting, it has been declining and it should be noted that it’s at it’s lowest percentile ever.  It will be interesting to see how Microsoft uses this influence to roll out Bing.

I really hate those market share numbers because most people aren’t given a choice or would use whatever they were told to or taught on.

Evolution

I have no intention of pissing off anyone, and I don’t have anything against anyone’s choice for their sexuality.  It’s their choice, and they should act however they feel.

I think the purpose of life is for it to continue (or one of it’s purposes anyway).  Species have evolved over time to adapt to survive in their environment.  I can’t think of a single instance where this adaptation was brought upon by the species’ choice… can you?  Some species can reproduce on their own, others require a mate of the opposite sex.  Humans require a mate of the opposite sex.  A human male cannot naturally give birth to offspring.  A human female cannot impregnate herself naturally.  So, how does homosexuality play a roll in the continuation of the human race?

Let’s assume the human race as we know it today gets “reset”.  All technology and science as we know it is lost.  We have to resort to living off of the earth.  What would become of those “choices” people have made regarding their lifestyles?  Would we, as humans, adapt?  If homosexuality is really engrained in DNA rather than a conscious choice, would it die off eventually?

Homosexual couples adopt, which I think is a great way to start a family.  I don’t really have any good arguments against that, except that in my “reset” scenario, I don’t think adoption would be a choice.  Adoption should not exist, IMO… I mean, it should not be needed.

There’s also surrogate birth, but I can’t see that playing into the “reset” scenario either.

Yeah, that “reset” scenario’s a long shot, but it could happen.  I try to break down every question as far as I can, because I usually find that the simplest answer is almost always the right one.

I don’t disagree with homosexuality because it disgusts me or it defies my beliefs.  I disagree with it because it’s not natural… it’s not life’s way.  I’m sure I’m wrong, so tell me all about it in the comments.

Apple Needs Microsoft?

I’ve debated this topic many times… does Apple need Microsoft to continue growth?

I’ll focus mainly on the workplace, because I feel that is the primary environment that has the most influence over what choice people make.

One could argue that all Microsoft would have to do is stop making Office for Mac, then the Apple OS in the workplace would diminish greatly.  Yes, there are alternatives such as iWork or OpenOffice.  Neither offer equal replacement functionality of MS Office, so deploying them in the corporate world wouldn’t work.  Installing Office for Windows in Crossover for Mac?  It’s buggy and lacking if you ask me.  Running a Windows virtual machine?  Works great, if you don’t mind everything slowing down quite a bit.  It also drives up your costs substantially.  I would get irate having to have a VM running all the time.  Also, could you imagine trying to train 100+ employees how to work back and forth between a VM?

One could also argue that MS Office is Microsoft’s cash cow, and their losses would be greater than their profits if they stopped making Office for Mac.  This would probably be short term because the previous Mac users would have to switch over to Windows, so the lost costs would eventually be made up.

I personally think, if Microsoft was fed up enough with the OS wars with Apple, they would be willing to sacrifice some short term revenue to end the competition once and for all.  So, maybe Apple should be rethinking their whole ad campaign strategy?

Then, of course, one has to consider what Apple’s primary consumer base is.  A lot of people have converted to Mac because of the flaming wreck that is Vista, mainly because the software they need to get their work done exists.  How many of these converts do you think would have made the jump if not for Office for Mac?

Is it possible that Apple would work harder at improving iWork’s compatibility with Office?  I could see them pulling from and improving on the OpenOffice code base.  Seamless integration would go a long way, but I have the feeling that most corporations wouldn’t be willing to risk dealing with the large amount of potential problems that would arise during this transition phase.

What do you think would happen?

P.S. – I’m proud of myself for not once referring to Microsoft as M$.  :)

More Microsoft Lies

Kevin Turner: Midmarket CIO Summit

Vista today, post-Service Pack 2, which is now in the marketplace, is the safest, most reliable OS we’ve ever built. It’s also the most secure OS on the planet, including Linux and open source and Apple Leopard. It’s the safest and most secure OS on the planet today.

I really, really, really, really hate crap like this for two reasons:

1. It’s an outright lie to improve sales.
2. It makes my job harder.

The first reason doesn’t bother me so much.  I can live with marketing FUD.  The second, however, bothers me quite a bit.  It puts the burden on me, an IT professional who has to support actual end users on a daily basis, to prove beyond a shadow of doubt that Windows IS NOT more secure than Linux or OS X.  In an academic/research environment, the person who is more credentialed is believed first.  So in this case, I’m up against a Microsoft CIO.  I can see the future conversation now:

• Them – “Hey, I read Windows is more secure than Linux or OS X”
• Me – “That’s a lie to improve sales on Microsoft’s end”
• Them – “But the guy who said it was Microsoft’s Chief Information Officer.  He must know what he’s talking about”
• Me – “You would think so, but he’s putting his company’s interest over his customers’.  It’s a lie to improve sales.”
• Them – “Are you saying you’re smarter than a Microsoft CIO?”
• Me – “No, I’m saying it’s a lie.  Windows is not more secure than OS X or Linux.”
• Them – “You expect me to believe you over a CIO?”
• Me – “In this case, yes.”

Then I go on to have to explain myself and cite references of proof, in which they will lose interest 5 minutes into my explanation.

So, because of a careless statement by someone who should know better, it’s effectively made my life harder.  Thanks Kevin!

Time Machine on OpenSolaris ZFS over iSCSI

I did it.  I finally got Time Machine working over iSCSI on my OpenSolaris RAIDZ1 ZFS pool:

For the most part, I followed the instructions here:

http://opensolaris.org/os/project/qosug/how-tos/zfs_iscsi_integration

Step #6 didn’t make sense to me (not sure what I was supposed to do), so I skipped it.

It also wasn’t quite as plug and play as the article suggests.  Here’s some of the problems I had to overcome:

1. My SCSI card wasn’t recognized by OpenSolaris.  This Google Group thread really helped.  Basically, I had to get the SUNWadp driver from a different Solaris distribution (I used SXCE).  I’m still not clear about how pkgadd works with a directory, so I just copied the uncompressed driver to /var/spool/pkg then ran pfexec pkgadd.  I also found (via dmesg after rebooting) that the driver is 32-bit only, so I had to remove $ISADIR from /rpool/boot/grub/menu.lst to boot in 32-bit mode.  I also found out that’s not really desirable with ZFS as it favors 64-bit but it’ll have to do until I can find a newer SCSI card.
2. I had my ZPOOL created, but kept getting errors when I tried to turn on iSCSI.  I then proceeded to try to install the COMSTAR iSCSI target pkg, which didn’t work.  I found out it was already installed and all I needed to do was enable the service.  Doh!  I’m still not clear on CLI-fu on OpenSolaris, so I used the Services GUI to do this.  I actually looked for the command to do it via CLI, but it seems to have changed quite a bit from release to release.  I was also getting impatient and wanted results!  :)  The good thing is, the GUI shows you the command, so it’s a good way to learn.

That’s it!  Thankfully once I had done that, I was able to connect to the target via the globalSAN iSCSI Initiator for OS X.  I actually expected it not to work, but it worked without a hitch!  I was able to format the ZVOL HFS+ so Time Machine could use it, and my backups began!  This was much faster performance-wise than AFP on Linux, and I didn’t have to trick time machine into using it!

Things I still need to do:

1. Figure out how to allocate the rest of the ZPOOL and share it via CIFS & hopefully AFP!  How do I tell ZFS to use the rest of the pool?  zfs create -s -V …
2. Get better with Solaris/OpenSolaris CLI.  Coming from Linux… it’s different.
3. Find a supported SCSI card and use 64-bit.
4. Figure out how to automatically mount/unmount the volume on my laptop for when I get to work or leave.  The trigger should be whenever I plug in my ethernet cable, but ONLY for that location.  Tricky…
5. Figure out iSCSI authentication, such as CHAP.

iSCSI uses a LOT of TCP overhead, so keep that in mind if you’re doing it yourself.  I have a gigabit switch in my office that all my machines hook into, so that traffic is kept off the main network.  I would not try it over wireless (if that’s even possible).

The more I learn Sun, the more I’m realizing it’s not really for the “latest and greatest” technologies, but rather a rock-solid stability.  OpenSolaris seems to be trying to bridge that gap, so I’m looking forward to seeing where it goes.