Hosting and the Cloud

Hello folks.

It really seems like an eternity since I updated this blog. I have been busy implementing some projects at work and Paul spent a fair chunk of time gallivanting around Europe and South America.  While I am sure that Paul has had a good chance brush up on his French and Spanish, I have spent a good chunk of my time brushing up on DFS logs and learning more about The Cloud.

Much has been written about the cloud and both the media and the big players in the IT industry seem to be having a virtual orgy espousing the virtues of server-less IT environments.I jumped in to this love fest at first too, but that was with great naivety – I was but a virgin in this game.

I know better now and let me tell you – don’t party in the cloud until you are sure you know what you are getting yourself into. While many of the benefits that big players like Microsoft and Google promote are true, they often fail to mention how difficult the actual logistics of moving into the cloud can be.  There are some real gotchas to watch out for. Put your party hats on boys.

Let’s take the example of a hosted exchange environment and dissect it a little. It seems like a great idea. The ability to take your email anywhere; no need for a VPN; 99.9 percent uptime and no menacing exchange limits or management to perform. All of this is true, but the question is, how do you get the data there?

Many organizations have spent years building up complex linkages in their exchange environment and to that end have enormous amounts of email behind all of that. There are public folders, contacts, resource mailboxes and a plethora of other oddities.  Quite simply, users don’t often realize how complex their environment is and will easily consent to changes to the environment without actually realizing what the change itself means. This leads to disappointment when the new environment is rolled out and a lot of headaches for IT support staff who were assured that everything was “kosher” before the transition.

Then, there is the logisitcs of actually moving data in the first place. The tools that exist for importing and exporting mail are quite labour intensive and demand that the user be pulled away from their computer. It can take literally hundreds of hours to export pst files, re-import them and then to have that data spool down again and rebuild a users cached mail file. In theory, there would seem to be many tools at one’s disposal for doing this, but at the end of the data these tools prove utterly unreliable for the large mail stores of today’s users. For instance, it is not implausible to run across mailboxes that are over 15 GBs in size today. How do you even manage getting such a mailbox into the cloud? How do you deal with exporting it. All of these operations actually require exporting mail to a .pst file. Exmerge is of little use since it breaks the files into 2Gb chunks. So, you are left with actually exporting this out of outlook and having to babysit it to ensure that it completes. Then you have to physically open the users new mailbox in outlook and import that data back in. Given average Internet speeds for small businesses, you are looking at a 4 or 5 day process just for one mailbox. The logistics are a nightmare. This definitely isn’t the no-strings-attached fun you were promised on Craigslist. 

Now, after all of that data is finally up in the cloud, what happens when this all breaks? You are completely at the mercy of some company in New York that has very little accountability to the end customer. I guarantee you that they will not feel the same sense of urgency that your own personal systems administrator feels when exchange has issues. Sure, it is more reliable – but let’s be honest. Everybody has some downtime.

Anyway, this was all a bit of a ramble, but I guess my point in all of this here is to make sure you carefully evaluate whether a hosted or int he cloud solution is really right for your business and if you do decide to take the plunge carefully audit and where possible purge the amount of data that will need to be uploaded. Then ensure that your users understand the transition plan 100 per cent and make sure that the expectations of the outcome are crystal clear. Do not promise that an environment will look identical to the in-house exchange. Just promise that there will be a way of accessing data that is needed in a timely fashion. Both your users and IT team will reap the rewards of this careful planning and management of expectations.

Thanks for the read!

Cheers….

Matt

My SBS CALS Have Disappeared

 

So here I ask, is there anything more intensely irritating than a boy dragging his nails across a dry dusty chalkboard? Probably not, but I must admit, every time I open up a licence manager on a SBS Server and look at CALS, I get that same feeling of impending rage.

Let’s just say it; SBS 2003 CALS were a total pain and completely useless.  Yes, I understand that Microsoft wanted to protect their best interests and needed to make sure that the SBS package wasn’t used inappropriately, but surely it did not warrant a system so maddening.

So, with my thoughts on this known, you can imagine how I felt when one of my client’s SBS Servers for whatever reason decided to drop its licence database.

This shouldn’t be such a big deal, except for the fact that nobody could find any records regarding these CALS and even calling Microsoft was futile. The previous IT firm did not document anything and it was not completely clear how they even purchased the CALS in the first place.

After discussing it with Microsoft and being told that they were “escalating” the issue, I set out to see if there was a technical solution to my problem.  I poked around and found that the licence database itself is implemented in two tiny files. These are: autolicstr.cpa and licstr.cpa.

Apparently, there is an automatic backup created of the licstr.cpa which is the main licence database.  The easiest way of fixing this is to copy these two files out of the C:\windows\system32 directory, use NTbackup and back them up to a file and then use the restore wizard in SBS licence manager to recover them from the file.

If you get an error message that the CALS are no longer useable, simply write down the keys that are shown in the licence manager, re-enter them and phone Microsoft to re-validate. Naturally, you should take the time to properly document the CALS and back them up using the backup licence function in the licence manager, so that you will never have to deal with this again.

This little trick sure made my day, and I hope it helps some of you as well!

Cheers!!!!

Outlook 2007 Inbox Renamed

We’ve been a little shy on updates lately folks. It has been a busy time of year, Paul is down in South America having the time of his life, and spring is fast approaching. Still none of this excuses our lack of writing and definitely doesn’t excuse the brevity of this post.

Nonetheless, this is all I have to offer for now.

A few days ago, one of my clients called me up and said that his Outlook 2007 inbox had suddenly been renamed to image001.jpg.  I have seen a lot of strange things, but this I really had to see.

As my log me in session connected and I focussed on the tiny text on my laptop screen, there it was. image001.jpg. This was indeed bizarre. I wasn’t really sure where to begin with this. The issue, in fact, doesn’t affect the end user in any way nor does it impede the normal flow of mail. But still, it is an annoyance. 

Historically, I have had great luck with the various switches that can be used to launch Outlook. So, I started digging in a bit, and sure enough I found a switch /resetfoldernames.

I was a little worried about running some random switch that I knew very little about, but again, past experience has been really good with these Outlook commands and away I went with it.  5 seconds later and the problem is gone.

So, for thos of you that have never used an outlook switch, here is the procedure. Open the run dialog box and enter the following command.

Outlook.exe /resetfoldernames

I always like to start this with outlook closed so that it opens one new instance of the program.

If anyone knows why this happens, I would love to hear about it. Please post your comments below.

Cheers!

Fixing RSS Feeds in Outlook 2010 Beta 2

I love RSS.  I can't believe I'm saying this, because for years, I never really 'got' it.  I knew how it worked technically, I saw how it could possibly be useful, but I just never really cared enough to actually use it, and I couldn't imagine why anyone else ever would, either.

That is, until Outlook added RSS support.

For me, Outlook is always running.  I manage half my life through that program, in fact (the other being OneNote, which I like even better).  And as I’ve mentioned before, I really dislike the current state of the web, where each web site is most like a program, with its own interface and quirks and learning curve, not to mention bugs, advertisements, distractions, and security issues.  So if I can take the content from the sites I love and read that content in a program I already know, I’m a happy man.

Outlook lets me do just that:

By default, Outlook creates a folder for each feed to keep things organized, but I’ve changed that to have everything delivered into my RSS Feeds folder.  Outlook keeps every article around until I choose to delete it, shows me which I’ve read, and lets me manage them just like any other Outlook item.  I can even forward interesting stories to friends just like an e-mail.  You can also see Outlook keep an archive of all our Slick IT articles here, which is a very easy way of having a complete backup and searchable offline reference of all our content.  Happily, the feeds I choose to read are saved in my .PST file, so I don’t have to re-add them every time I rebuild my machine or reinstall Outlook.

The Problem

When I upgraded to Outlook 2010 Beta 2, however, the magic stopped.  Everything looked okay… my RSS Feeds folder was still there, the articles I hadn’t yet read were working just fine, but I wouldn’t get any new articles delivered.  It wouldn’t sync.  I’d hit Send / Receive all day, but the RSS Feeds just wouldn’t update.  I even checked under Accounts, where they seem to show up:

But despite this, Outlook wouldn’t actually check them.

The Solution

It turns out to be quite an easy fix.  Here’s what worked for me:

1. Using Internet Explorer, add a new feed.  If you’re using IE now, just click the View Feeds button on the toolbar and then click “Subscribe to this feed”:
   
   
2. Under the File menu, choose Open, and then click Import:
   
   
3. Choose “Import RSS Feeds from the Common Feed List” and click Next:
   
   
4. Check the feed you just added and click Next:
   
   
5. Click finish, and then try another Send / Receive.  You should find all your old feeds now begin to synchronize.  You’ll only have to do this once, and you can now remove the feed you just added.  But, if you just added Slick IT, why would you?

Hope this helps!

The Fall of DomainPeople

Like many of you, I manage a few domains.  Mostly, these are on behalf of clients, and as such, great support and reliability are more important to me than rock-bottom prices.  I want to keep all my domains with the same registrar, under one account, and since many of my domains are .ca, this somewhat limits my choice in registrars.

Well, let me summarize this lengthy article with a few choice words of wisdom:

Never use DomainPeople for anything.  If you’re using them now, run away.

For years, I’ve used DomainPeople.  They have been around forever, they promise 24/7 tech support that’s easy to reach (no waiting on hold), and they offer a good selection of “value-added” services, such as DNS hosting, e-mail forwarding, and so on.  Their prices might not be the cheapest, but they’re not quite as astronomical as a couple others you probably know about.  And, keep in mind, this goes back a few years, to a time when GoDaddy wasn’t an option, Network Solutions didn’t do .ca domains (or perhaps was even more expensive than it is today), and generally, the world of domain registration was a very different place.

By the time 2009 rolled around, DomainPeople’s site and control panel was looking a bit dated.  There had been essentially no change for years in either the function or the design of their services.  But that’s okay.  It all worked.  Sure, they weren’t the most dynamic and aggressively-growing company out there, but I didn’t want that.  I just wanted a good, stable registrar.  I didn’t mind putting up with a quirky interface from the 90’s so long as they delivered where it counts.  And for years, they did.  Mostly.  And when they didn’t, it was easy to reach them by phone and have everything sorted out.

Until everything changed.

The Full Story

Let’s do this chronologically.

January 11, 2010: The first sign of change arrives.  I receive an e-mail (well, five, actually… but DomainPeople has never been particularly organized when it comes to that sort of thing) describing some upcoming changes to their control panel.  Here’s essentially what they had to say:

• Within two weeks, your account will be automatically upgraded.
• You will gain access to new features.
• Your username and password will not change.
• An auto-renew feature has been added.  If you choose to keep a card on file with us, we’ll renew any domain automatically before it expires.
• You will no longer be able to access the control panel on a per-domain basis using the domain’s password.  If you don’t want all your domains to be accessible by your master account password only, you must remove them from your master account within the next two days.

Okay, so fair enough.  New features are coming, and there’s nothing I need to do.  Wonderful!  Now, in the back of my mind, I was thinking “gee, it’s a good thing I don’t need to provide my clients with access to their domain control panel, because then I’d have only two days to set up new accounts for each and every domain, provide my clients with the new login information, and then lose the ability to manage the domains myself without keeping track of dozens of user accounts, hoping none of my clients changed their passwords”.  This should have set off a few alarms… I’m sure many of their customers were caught off-guard by this, and I doubt many of them actually responded to this in the two days DomainPeople gave.  But it didn’t affect me, there was nothing I needed to do, and I was looking forward to having a snazzy new interface with new features, so I didn’t much care.

Friday, January 15, 2010: I receive close to 10 e-mails from DomainPeople, all of which are notices of pending auto-renewals.  I also receive two notices of auto-renewal failures.  This is all a bit odd, but I figure that since I haven’t provided a credit card, this is just their auto-renewal system trying its best to do its thing.  It’s a bit odd that I was given no notice of pending auto-renewal for the two failed attemps, but that’s okay.  So far, none of this means anything; it’s just a bit of weird e-mail related to a feature I’m not using.

Saturday, January 16, 2010: I receive two notices of pending domain expiry.  This is weird… these are different domains.  I would have expected more auto-renewal notices.  So do domains renew automatically, or not?  Still, I put this down to glitches in the deployment of their new system.  Again, none of it affected me at the moment.

Sunday, January 17, 2010: Another five notices of auto-renewal failures.  This is getting a bit ridiculous.  By the way, at this time I was on a small island off the west coast, and due to a recent windstorm, had no power or Internet.  But this is all okay, because I don’t need to do anything.  I thought.

Monday, January 18, 2010: The calls start.  I’m not quite sure why so many of my clients have no e-mail or websites working at the moment, but my first thought is that it must be related to the power outage.  What weird, vital, forgotten thing do I have running through my workstation?  It’s tough to diagnose, of course, when all you have is a cell phone (thank God for Windows Mobile, and no, there is no iPhone App for that), but by that afternoon, I was eventually able to trace things back to a DNS issue.  Well, for the clients having trouble, I use DomainPeople for their DNS.  So, I call them.

They answer quickly.  This is a good sign.  I explain the problem, and they admit that yes, there was an issue transferring zone files to their new system, but there are technicians working on it right now, and the problem should be fixed in a couple hours.  So, I leave them to do their thing.  The notices of renewal failures and upcoming renewals continue.

Tuesday, January 19, 2010: Still no fix.  And still no power, unfortunately.  So, I call again.  I’m given pretty much the same story.  They’re aware of the issue, they’re working on it, and it should be fixed soon.  I’m also asked for a list of ‘high priority’ domains, which they promise to take care of right away.  After asking a few more questions, I find out that these issues would have started on Friday, and yes, many other customers are affected.  I’m a bit discouraged at this point… four days of downtime on a simple DNS server, and no estimated time of resolution?  This is not the DomainPeople I had grown to love and trust.  Still, shit happens, and it sounds like they’re doing everything they can to fix it, so I hang up feeling pretty lucky I’m not a DomainPeople admin today.

By the end of that day, I realize how much as gone wrong: no DNS for any of the domains they serve, no e-mail forwarding, nothing.  Web sites are down, e-mail is broken, and still, there’s no estimated time of resolution.

Wednesday, January 20, 2010: Still broken.  And another surprise: a bunch of my domains have auto-renewed!  How odd. I call tech support, and learn that there is a senior technician working on the issue, but this person will not accept my call or provide any further details to Tier 1 tech support.  No estimated time of resolution is available.  Apparently, I should just hang up, wait, and eventually, it will be fixed.

I inquire about the domain auto-renewals.  Well, it turns out that one of my clients had called DomainPeople directly, and was told the issue was caused by a domain name about to expire (this was entirely false), and if she didn’t renew the domain immediately, it would be gone forever.  In a panic, my client provided her credit card to renew the affected domain name.  Big mistake.  This didn’t fix the problem, of course, but it did allow the auto-renewal system to renew many domain names to her credit card.  The tech support representative was not sympathetic, but did suggest I call their billing department.

Billing agreed with my assessment, but didn’t see anything wrong.  I was told that the auto-renew process was explained in an e-mail, and that the system was working properly.  I should have turned off auto-renew if I didn’t want this to happen.  Well, here are the problems with that:

• Notice of pending auto-renewals were first given only on Friday, the 15th.  This is, by the way, the same day everything went down and the first auto-renewal failure notices started to come in.  No credit card was on file at this time.
• DomainPeople did not have authorization from me – the accountholder and domain registrant – to make changes to the billing configuration of my account.
• DomainPeople did not have authorization from the cardholder to charge the credit card for the amounts or services billed.

Eventually, the billing representative was able to see this.  The new system, however, does not have the ability to remove a credit card that has been added, and the charges could not be removed, but only reversed and then applied to a different credit card.

So, at this point, DomainPeople has told my client the downtime was due to an expired domain (entirely untrue), has charged my client (without authorization) with the renewal of many other domains (none of which were immediately about to expire), and has still not fixed the technical issues.  No one is able to offer any additional help.

I then track down a manager at DomainPeople who listens patiently to my tale of woe, takes a bunch of information, and promises to have both the technical and billing issues resolved as soon as he can make it happen.  He promises to stay in touch with me, and tells me to expect a call from him the next morning.

Thursday, January 21, 2010: Still broken.  And no call from tech support or the manager I was working with.  I do, however, have a message from the billing department advising me that auto-renew has been turned off, they’re working on the refund, and I should call to provide a bit more information.  I do, and a told that they will be able to process the reversal in the next day or so.

However, the technical issues still exist.  And worse, the tech support department has no record of anything even being wrong.  Nobody is working on this.  The senior technician responsible for this problem is ‘not available’.  The manager I spoke with earlier is ‘not in today’.  And nobody knows anything.  I’m encouraged to open a ticket using their web site.

Again, I battle my way to a senior staff member who seems to be in a position to help things.  He is very friendly, understanding, and promises to get things resolved immediately.  He takes my contact information, a list of the domains in question, the support ticket numbers I have been provided, and other details.  He then tells me he will call back as soon as he’s able to get things resolved.

Friday, January 22, 2010: Still broken.  I’ve heard back from billing, and they are pleased to say they were able to reverse the charges.

But not a word from tech support or either of the higher level people I was working with.  I’m told to sit tight, that technicians are working on this, and I will be called.

Saturday, January 23, 2010: Still broken.  Still no word.  I’m out of town and unable to do much except for call to see what’s happening.  I don’t learn much.  This marks one week since my clients had working e-mail.  I do not look good to them.

Sunday, January 24, 2010: Still broken.  Still no word.  Another call, but no more answers.  Only empty promises.  Nobody I’ve spoken with earlier is around, and nobody else seems to be in a position to help.

Monday, January 25, 2010: Still broken.  Still no word.  Things are getting really critical at this point.  I’m starting to ask myself how I’ve let things go over a week.  I could have transferred the domain to a different registrar twice over by now, at least.

I call tech support.  No changes have been made.  They have no record of any word being done on this.  The tickets I provide have all been marked as resolved, one by the original senior-level technical working on my account.  This technician still refuses to speak with me directly, and does not return the messages I have left for him.  Neither of the two managers I spoke with are available, nor have they returned my calls.  I leave additional messages, explaining how I cannot wait any longer, and this issue, one way or another, must be resolved.  I am yet again promised immediate action.

I have also now come up with a strategy on moving away from DomainPeople.  It’s not pretty, or fast, but it will work.

Tuesday, January 26, 2010: Still broken.  Still no word.  At this point, I’m done with DomainPeople.  It’s been over 10 days since this problem arose, and I am no closer to fixing anything than when I started.  I’m going to reconfigure my DNS elsewhere and point my domains to these new nameservers.  I’ll also reconfigure e-mail forwarding, avoiding DomainPeople’s services and servers entirely.  This will all take time, of course, but a couple more days of downtime followed by a guaranteed resolution that I can control looks much better than anything DomainPeople could offer at this point.  There’s only one thing that could break this: in the unlikely event that someone at DomainPeople is still working on this, they might see the changes I make to my nameservers and undo them.

So, I call.  Nobody knows anything about my issues, and nobody cares.  All tickets are closed.  There is no one able to help.  Nothing.

I leave a message, stating that no one at DomainPeople must edit my nameservers, or do anything else with any of my domains, without first talking with me, directly, over the phone.  I’m assured this will be the case, but I don’t think the representative actually does anything.

I begin the process of rebuilding DNS records elsewhere.

Wednesday, January 27, 2010: Still broken.  Still no word.  Everything on my end is now set to go, though, and I’m just waiting for the nameserver changes to propagate.

Thursday, January 28, 2010:  All my domains are now working!  This is because I moved all services away from DomainPeople, though… still no word from them, of course.

The Aftermath

If I look back, here’s what I see: for many, many years, I have a great relationship with DomainPeople.  I pay them thousands of dollars, they provide me with everything I need to offer quick and reliable service to my clients.  And then one day, the world explodes.  With no useful warning whatsoever, they break half my domains, charge one of my clients’ credit cards with a long series of unauthorized charges, and offer no assistance whatsoever in fixing anything.  For over a week, they do nothing to fix the problems they have created.  They do not return calls.  Ultimately, I am held hostage by a company who doesn’t care and doesn’t respond, and I am left to my own devices to find a way to escape and restore service to my clients.

It’s important to realize that I was never rude or hostile with anyone at DomainPeople.  Every time I called, it was always easy to find someone to speak with.  When I spoke with managers, they were all very understanding and promised to do everything they could to resolve the matter.  And while all this sounds good, the fact remains: they didn’t fix what they broke.

Hosting domains and running DNS is not hard.  These are the fundamental parts of the Internet designed to withstand nuclear assault.  There’s just about nothing that needs to be done here.  And there’s absolutely no reason why this should have broken.  That it did, and for so long, is enough to stay away from DomainPeople alone.

But to have a large,  legitimate company (who charges premium rates for their services in exchange for promised reliability and support) completely ignore a customer like this and not work towards resolving these issues is entirely unacceptable.

So, everyone: be warned!  Go to great lengths to avoid DomainPeople, because you do not want to go through what I did to learn this.

And, of course, if anyone has a recommendation for a new registrar that does .CA domains, offers DNS hosting and e-mail forwarding, and actually helps out when there’s a problem, please leave a comment!