Category Archives: Extreme Availability

Category for our extreme availability blog topics and posts. Look here for thoughts and advice on the subject.

The Extreme Availability Blog is our attempt at framing what this is and is not, and establishing a solid dialog on the subject in the hopes of raising the expectations we have for our friend the computer. In our daily experience with computers, we see spam, viruses, slow web pages, crashes, spinning clocks, and other minor annoyances that really don’t help our perception of what computers can do. What we don’t see is the infrastructure that quietly runs in the background, making sure that our money is moving around correctly without prying eyes, running our power plants, giving us the security of knowing that we can pick up the phone and call 9-1-1, and letting us go to the grocery store and pay for food with confidence that the computers will be up, even if our credit is maxed-out.

Auto-Tune R1.6 is in Progress

We are currently putting the finishing touches on Auto-Tune R1.6. When it is released, we will give you one month of free time and updates so that you can enjoy trying the new C++ capabilities.

Biggest news in this release? Dynamic SQL support for C and C++. You will be able to scan some of your strings for SQL statements and get plans without having to cut and paste into mxci.

Welcome to 2015 NonStop Customers and Partners

As we look at 2015, we cannot help but see x86 chips in the crystal bowl. And those will be a really interesting challenge for us all. How are we going to refit all of our development, QA, and production systems with copies of our new infrastructure? Or are we? It turns out that virtually all of your current native Itanium code will build and run on the x86 platform. That is good, right? Absolutely. It means you can keep developing at a breakneck pace – no waiting, no stalling, this is NonStop people – and then when your x86 machines show up, switch the build settings in ECLIPSE over to the new target and get right into integration testing. But wait, what if you are not in ECLIPSE yet?

Fortunately, you have some pretty awesome partners out there to help with this effort, including us. This includes people who know how to migrate off of code 100 compilers to workstation development including, yes, BASE24. We know, because we are one of them. We have been building ECLIPSE-based solutions for NonStop since 2004.

But the question came up a few months back: Whyever did you pick NonStop, of all platforms? Interesting but long story. A shorter one would be: Why did you stay with NonStop? That is a good one. Thanks for asking. NonStop represents the Massively Parallel Processing (MPP) a class of systems that is the future of Big Data. Without architectures like ours, we could not possibly hope to deal with the imminent BrontoBytes of data that are headed our way. The multi-core per logical processor structure further affords a strong measure of reliability for extended-mission-critical applications. Most imporant to us is the strong operational discipline that exists on NonStop for always-available management of which other platforms only dream. That means Bigger, Stronger, Faster! But there are really only three critical success factors to keeping systems running: Process, process, and process. That is why we are here.

Savings comes from moving to workstation development.

Process is why we do what we do. Nexbridge started with a strong process focus, and our original services reflected that. We stayed with NonStop and build products around a significant gap in sound processes, that still exists – how do you build large-scale reliable production hardened applications using (now 2015) technology while reducing price-points? Everyone wants to save money, and that savings comes from moving to workstation development. Yet, the risks from inappropriate source-code visibility to deployment errors to simply a lack of technology presents challenges to NonStop. Solving that is also why we are here.

Whether you need products or OSS and workstation development services, we are here.

By adding SQL/MX Auto-Tune to our ECLIPSE suite, we have made SQL performance evaluation simple for C, C++, and Java developers, architects, and managers. Use it to evaluate how plans change over time or between different database; compare statement efficiencies; and try different performance options without having to modify your code.

Don’t forget our ECLIPSE Pathway Plug-In, which provides definition, management, and monitor of Pathway subsystems.

And for those heavily into the original NonStop DDL, our ECLIPSE DDL compiler allows you to preserve your investment and move into workstation development quickly and efficiently.

Please take a look at our product portfolio and the latest article in Connect on What Git Means to the NonStop Community.

Auto-Tune R1.4 Released

Welcome to 2015! Release 1.4 of the SQL/MX Auto-Tune plug-in has come out of our build machine and is now on the update site. We had to make a significant change to the plug-in extension point for this release, but since there are no production extenders, at this point anyway, the impact should be minimal. Please let us know if this is an issue and we can provide a backward compatible version. Note: We normally do not make incompatible interface changes, but this was a functional omission that needed to be corrected.

Connection sharing and profile management are significantly improved with this release and people are encouraged to upgrade.

When Availiability Meets Legislation

Well, it looks like Canada is about to get an interesting kick in the business behind from government when it comes to business continuity. What? How could that be, Randall? What does legislation have to do with my ability to continue operations? Good question. You may not like the answer.

The new legislation is considered the most aggressive in any jurisdiction and has some provisions different from any other country on the planet. In the “Fast Facts” section of the new law, coming into effect in July 2014, it will generally prohibit:

sending of commercial electronic messages without the recipient’s consent (permission), including messages to email addresses and social networking accounts, and text messages sent to a cell phone;

Now, this looks really good on paper, until you dig deeper into the legislation. Suppose you have a customer that has installed your amazing banking software in production. Unless you have specific permission to contact your own customer by email, and you have to get it periodically because it expires, you may not be able to send them information about critical updates (fortunately not including mandated PCI security fixes and product safety notices). Out of warranty operational issues are not covered. I’m sure they won’t mind hearing about it, but you still have to get their permission. No problem, right? Well, suppose that you have done a great job and there have been no fixes or bugs or any other communication for two years. Their permission is out of date. You can’t tell them that something is wrong. Not by email anyway. What are you going to do? More importantly, what are they going to do? They get hit by the bug you fixed, but you couldn’t tell them. They are down and you might get sued, or not, depending on where province/state/country your license agreement is governed. Oops.

You cannot even send an email to a prospect asking them for permission to get in touch with them in most cases. If you had doubts about email being dead in Canada, doubt no longer.

The way this could work is to have a mechanism similar to Add Friend buttons in social media. If the legislators had worked with technology companies, this might have worked. It was discussed more than a decade ago. However, this was not done, so we now have gaps between legislation and technology that will result is a fair amount of chaos and damage to the economy.

On the other hand, the entire email spam situation can simply be implemented, without any legislation, if people use White Lists as the only allowable means of email receipt. Using that choice, no one would get spam or any communications whatsoever from anyone new, which appears to be the point of that part of the legislation.

There are a few things you can do, as a customer, to help your supplier operationally:

  • Consider “out of warranty warranties”. This would be a provision in your license agreement to have you be notified of all changes as warranty items, even if you are not paying for (or entitled to) warranty service.
  • More importantly, if there is any doubt, make sure you contact your vendor periodically to keep implied consent active. Or make your consent explicit.

As a vendor, make sure you have consent on record. Regardless of what may have been said in the press, and there are exceptions, implied consent essentially expires every two years without any other communications. So what you can do:

  • Keep in touch with all of your customers. This will keep implied consent alive.
  • Get explicit consent.

We’re still researching the proper procedure in dealing with this rather painful situation. Needless to say, we may have to send out some bulk mailings to obtain your permission so that we can let you know the answer, at least if we don’t have one before this law goes into effect.

Unfortunately for the pain and effort that we legitimate companies will have to go to, the legislation is unlikely to make any difference to the nasty spammers out there, who will simply move to a different country and send us spam anyway.

All I can say here is, please keep in touch. Unless you are a spammer, we are not going to have a problem with you sending us email communications or contact requests. You are our valued customers, and will we treat you with the respect you deserve.

Surviving a Lengthy Electrical Power Outage

Winter brings an interesting set of challenges to business continuity. Temperatures drop and infrastructure gets covered in ice and snow. Maintenance crews have a much harder time gaining access to things in need of repair, including wires, underground transformers, and conduits. Power and drainage problems happen almost predictably every year. It’s not about if, it’s about when and how often.

If your business has been interrupted due to lengthy blackout this winter, or if you are justifiably worried about it, one of things you might consider is installing a backup generator.  These generators can be powered by either bottled gas (propane) or natural gas.  While diesel and gasoline powered generates are available, and usually slightly less expensive, you might want to avoid those due to the fact that the local gas station will be impacted by the same blackout unless you are prepared to store sufficient quantities of the fuel – and that means you may need permits to store more than a few minutes worth. There is value in those, don’t get me wrong. They’ll let you do a clean shut-down of your operation, but you should not depend on them for running your business during a lengthy outage.

Don’t be tempted to hook up a standby generator to your electrical system during an outage, if the power comes back on you can be electrocuted, destroy the generator or burn out any equipment that is hooked up. If you use a standby generator run it outside due to carbon monoxide and run a sufficient size extension cord directly to any equipment you need powered.

The next step is to determine the size of the generator you need. In a home you would take into account things like refrigerators and freezers, heating / cooling, sump or well pumps, internet, television / radio / phones, lighting and anything else you would like powered during a lengthy outage. At your office, think about what systems need to stay running – if you are running a factory, remember to include not only the power tools, but health and safety equipment.

These standby generators can be hooked up to the electrical system with an automatic transfer switch to turn off the utility power while the generator is running.  They have battery start and can have automatic testing on a weekly basis – a crucial feature in our opinion otherwise you might find you have just a very expensive decoration.  Natural gas and propane generators are manufactured by companies like GE, Generac, Cummins, Kohler, and Winco. The smallest units start around $2000 without installation for a 7000 watt unit. Some propane generators may be installed inside your well ventilated factory, but always check the manuals and regulations in your area first.

Stand-by generators generally will start running within minutes of the power going out – they take time to get going and stabilize – so you may still need UPS power for critical items, like servers and phone systems, that need continuous power. There will also be a short interruption when the utility power returns, which UPS devices will also handle.

Assess your needs and know how much power you need and when. Don’t get caught wishing that you were the one who bought that last generator instead of the person in line ahead of you who did.


Where Hockey Meets Continuity

After watching the Canada/USA women’s hockey gold medal match, I was struck by a rather strange thought. Business and Government had come to a temporary stand-still, while the whole country watched an incredible comeback. We cheered, we roared, we swore at the TV a lot, but other than looking down at email, what was going on? Nothing. The government even postponed scheduled press conferences until overtime was done. How Canadian, eh?

What does this have to do with anything? In Canada, hockey comes first, doesn’t it? Well, no. Electronic commerce comes all the time, whether you are looking at it or not. And that means that you have to be aware of what’s going on. So how do we manage continuity when the entire country is culturally predisposed to putting Hockey first? That’s a good question, but it is part of a larger problem.

How do we handle situations where people’s emotions and desires come before business? We adapt, that’s how. As managers, we have to be aware of what influences our staff. If we don’t or won’t, then we are going to have unexpected situations, like everyone disappearing to go downstairs to the local pub to watch the game. You simply cannot force people to act against their nature. You have to plan for it. So, when there is a gold medal hockey game scheduled, don’t plan on doing a massive benchmark or critical installation or code activation requiring everyone’s attention. Plan around it, because you know it’s going to happen. And ignoring what you know will happen will bring forth the biggest risk to your company: you.

When The Big Thaw Hits

As we watch the sun disappear behind the clouds today, we have to consider emergency preparedness and what happens when a rapid thaw occurs after there is a lot of snow.

There are two big questions:

  • What does this mean to me; I mean, I’m in an office, right?
  • What do I have to do to prepare, since, ok, I get it now?

So what does it mean? Rapid rises in temperature, in the winter anyway, result from weather systems. If you are a weather geek, you’ll know terms like frontal boundaries, occluded fronts, moisture pumps, omega blocks, and this years favourite scary term, “The polar vortex”. Huh? Ok, sudden warm weather means rain and melting. Usually a lot of it. We are currently forecast to get 30mm of rain, but we’re also going to get almost that in melt. So, Randall, that’s only 6cm of water. Who cares? Well, there are a few big problems for you. Let’s start at home.

The most important thing to do when you get home tonight (or now if you are home) is to clear your storm drains. If you don’t, water will simply flood the streets. But 60mm translates to a whole lot more because all that water in your yard has to go somewhere. We’re now up to triple that on the road, and that’s probably above your front-end air dam on your car. But wait, there’s more. The water is also going to flood into the buried power transformers, which do not like getting wet. Oh, and their drains are probably all frozen too. Worse, the salt on your street is going to mix with the water to make a really good conductor, so you need to make sure to avoid going near ground transformers during the storms and until the water drains off.

So the impact at work is, expect power failures and surges. There will be flooding on the roads, so you might want to consider public transit. Do you park in an underground garage? Did you remember to test your UPS batteries?

At home, clean those drains, have candles, and stay away from transformers. Monitor the situation and if you see water pooling up against the side of the house, get out the shovels. In fact, go out now and look for low spots in the snow. And if you’re me, check the UPS batteries at home also. They’ve had quite the workout this season.

You should also monitor the situation as it happens. The Weather Network has published a severe warning here:

You can also read my interview with The Richmond Hill Liberal by Kim Zarzour here:

Noah’s Flood

What will make things really nasty is the potential for the following: The high snow banks can temporarily keep some of the water in large pools. Eventually, the snow/ice mix will melt and have a catastrophic failure. This means a rapid break, rather like a dam bursting is possible in places. It is absolutely crucial that you and your children stay away from flowing water and avoid being below these dams. Safety first. For example, avoid the Upper Mill Pond park area during and immediately after this event.

For Geeks Only

If you are really interested in doing the math (If your name is Sheldon, for instance, and are reading this), take the ground topography of the area for your volume calculation. Assume a 1:3 road to property ratio, and that the weeping tiles are frozen solid so have no absorption at this time of year so do not factor into capacity. If you want to make things deeper, assume that all snow/ice on roofs, roughly 1:1 with the roads, are going to completely melt off and not pool on the properly, and that these will translate from 15cm snow/ice mix to 5cm of water (it would roughly be 1.5cm of water if it was just snow).  As a SWAG, this results in roughly 200-300cm of water that could pool without proper road drainage. But don’t hold me to these numbers. They are for planning guidance only. The number could be vary by a magnitude either way.

Dictionary Attacks

Hi Everyone. We’re currently the target of a broad-based dictionary attack from Bots in Hong Kong and the United States, trying to crack our administrator account, as evidenced in our log files. Fortunately, our defence measures are holding and we are monitoring the situation closely. If you encounter any issues on our website, please let us know.

The ISP administrators at the source of these attacks have disabled their own support/abuse emails, so there is no way to stop them at the source. Whether or not this is deliberate is subject of speculation.

Wisdom from WCDM: What Makes for Great Crisis Management Teams

I don’t normally cross-post, but this article is really a good bit of wisdom from my colleagues at the World Conference for Disaster Management, where I spoke a few years ago. The article is definitely worth the read, so I am recommending it.

This applies to large companies, governments, emergency managers, as well as a number of our clients who need the responsiveness and resiliency of being Indestructible. Please, if you read nothing else this week, read this article.


Accidental Blacklisting

This one was rather annoying, so I thought I’d share lessons learned. Yesterday morning, our web host provider had a number of servers accidentally added to BlackList 1. Not good, because it caused our outgoing email under the .com TLD to be blocked. This meant that all outgoing emails bounced. Fortunately, our backup .ca TLD domain is hosted elsewhere. We’ve switched over temporarily, and while annoying, it does show that unless a company has a backup for email, they can be shut down rapidly and with no notice. We are still waiting for resolution.