Emergency Maintenace Tonight

We will be performing emergency maintenance tonight on one of our nodes at 12 AM EST. This is only expected to affect a small number of customers and you will already have been notified directly if you are on that list. The maintenance is expected to last 20-30 minutes.

Deploying Builds to Your FAST Production Server

We previously discussed deploying new releases to your FAST staging server. More and more customers are now asking how to do their own production deployments. That’s exactly what we wanted to hear! To tell you honestly, production deployments don’t differ much other than the fact you’re dealing with live customers. Always take precaution in this scenario and keep in mind the handy cap deploy:rollback task that can immediately roll your system back to the previous version.

The first thing you can do is throw away your deployment file and use the one below:

Modified deployment file for FAST staging and production servers

Make sure you save it as deploy.rb in your config folder. Inside this file you’ll need to modify lines 2 and 3 with your appropriate SVN login credentials. Between lines 20 and 37 you’ll need to replace YOUR_PRODUCTION_SERVER_ADDRESS, YOUR_SVN_BRANCH_NAME, and YOUR_STAGING_SERVER_ADDRESS with the appropriate values. Keep in mind that YOUR_SVN_BRANCH_NAME is referenced on line 29 and 37. If you don’t know your SVN repo information then please contact our support team ((JavaScript must be enabled to view this email address)).

Now that you have your deployment file in place you can issue the following command to deploy to staging:

cap staging deploy

and the following to deploy to production:

cap production deploy

If you run into any deployment problems please be sure to search the application forum and post your question there.

Everything coming back online

Systems are coming back online now, we should have all production systems up very soon.

Problems continue

We can finally see a light at the end of the tunnel. We have access to Virtual Center again and connectivity to the VMs.

Unfortunately, at the same time we had very strange issues with our ticketing system (Jira). Some of you may have noticed comments that were meant for other customer coming to you. Jira runs on the same cluster so this may have been related somehow. In any case, a restart of Jira seems to have remedied the situation. We will be investigating what happened here very closely and taking steps to insure it does not happen again.

Cluster Problems Continue

We are still experiencing problems with the FAST cluster. We are starting to get some VMs to come back up, but many still remain problematic. We will continue to provide updates and apologize for the inconvenience.

FAST VMs on Node 3

We’re currently experiencing an issue with Node 3 of the FAST cluster. We’re looking into this issue and should have it resolved shortly. This will only impact a small group of our customers, but if you’re having problems accessing your system don’t hesitate to open a ticket by sending an email to (JavaScript must be enabled to view this email address). Stay tuned for more updates and we apologize in advance for any inconvenience.

New FAST Core Tracking Build #10340

We promised to get our app releases to happen more frequently so here you go. Just 30 days after our last release we’re publishing the latest core tracking build. We consider it a fairly minor release with a ton of enhancements from a build and deployment perspective. This should make it easier for developers to deploy applications to their staging and production servers. There are also some nice enhancements to the code. Here’s the diff file:

http://dev.numerex.com/uploads/files/fast_tracking_10340.diff_.zip

and a list of what’s new:

- Moved builds from Cruise Control to Bamboo. This will allow customer’s to view and manage their own builds.

- Moved to a new SVN repo which provides tighter integration with our build system and simplifies deployments.

- Fix for geofence centering in IE8 (http://code.google.com/p/gmaps-api-issues/issues/detail?id=1055)

- Support for ci_reporter and rcov in the new build system (Bamboo) to track build success/failures

- Reverse geocoder enhancements

- Support for NM-5500 (wired Xirgo) device in admin

- Notifier fix so that it can be successfully restarted via Capistrano deployments

- Rails app no longer runs in slicehost environments for both staging and production. It now currently runs in the proper environment. This means we can get back to the proper deployment commands: cap staging deploy and cap prodcution deploy. Yay!!!!!!

We’re getting heads down again for the next release so always feel free to ask questions in our forums and make suggestions via the wishlist forum. Enjoy!

Issue with FAST Deployments and Timezone Synchronization

We recently ran into an issue where our SVN server’s timezone was out of sync with our development environment. This was causing problems with tagging builds for staging and production deployment. While we’ve corrected the issue we’ve yet to announce our latest release, so we’re delivering this fix in the form of a patch file. It’s a minor patch that only affects one file (lib/tasks/deployment.rake). To patch your local project simply issue the following command from your Rails directory:

patch -p0 -i /path/to/timezone_deploy_10067.diff

For more information on patching please refer to the following article: Updating Your FAST Application with the Latest Release.

You can download the patch file here. Unzip, apply, and be sure to commit your changes to SVN.

Simplifying FAST Deployments on Your Staging Server

We’re currently in the process of improving our build system and overall deployment process for customers. In saying that we’re publishing two modified files for use ONLY ON YOUR STAGING SERVER. We realize that deployments have been a hassle and require some “tribal knowledge” about our systems. With that being said, please check our blog frequently as we’ll be updating you on our progress. The improved process means that you’ll have direct access to your build system as well as a much easier way to get deployments out to your staging and production servers.

These two files that will make staging deployments a breeze (we hope). You’ll need to place these files in your Rails project, but we don’t recommend checking them into your FAST code repository. Save them locally and you can always come back here to download them. They will become obsolete in the coming weeks and we’ll provide you with proper deployment files for both your staging and production servers. The first file is a modified deploy.rb file that you need to place in your Rails config directory. Simply rename your existing deploy.rb and place the following in its place. Make sure to remove the .txt extension.

Modified Rails Deployment File

Inside this file you’ll need to modify lines 1 and 2 with your appropriate SVN login credentials. You’ll also need to modify lines 20-22 with your appropriate FAST subdomain and SVN branch name. If you don’t know what these are then you should contact our support team ((JavaScript must be enabled to view this email address)). The other file you’ll need to replace is a modified database_procs.rb file that also lives in your Rails config directory. Just download, remove the .txt extension, and there’s no need to modify anything.

Modified Database Procedures File

With these two files in place, and the changes to deploy.rb mentioned above, you’ll be able to issue the following command to deploy to your staging sever:

cap staging deploy

You’ll be prompted for your FAST server password, which you can find in your welcome email. That should do the trick and please feel free to post in the dev forums or contact support if you have any problems.

Latest FAST Build #9502 in GitHub

As mentioned in a previous post the latest FAST build is ready for you to grab and tweak to your heart’s content. It’s publicly available from our GitHub page:

http://github.com/numerex/FAST_Core_Tracking

Enjoy!

New FAST Core Tracking App Build #9502 Released

The dust has finally settled and we’ve released the next stable version of our FAST tracking application. Version #9502. Here’s a diff file in case you want the juicy details:

http://dev.numerex.com/uploads/files/fast_tracking_9502.diff_.zip

Here’s what you’ll find in this release:

- Including mileage statistics on trip reports

- Trip report exports include mileage as well

- Expose all event types in the “readings” report

- Display “GPS Not Available” where appropriate and ensuring only valid GPS readings are displayed

- Support for an improved Numerex reverse geocoder (although still in testing)

- Fix to is_admin flag being lost when editing users

- New commands table in prep for a FAST_COMMANDs engine (next release)

- Geofence events don’t overwrite existing event types

- Improvements to how trips are detected

- Fix for issues with duplicate idle events

- Exposing gateway name in device admin page

- Support for handling out of order geofence readings

- Admin paging for accounts, devices, users, and Xirgo gateway

- Admin filtering for devices based on profile

- Display MSISDN in device admin when requested

As always, if there’s anything you want to suggest for new features then please post in our wishlist forum: http://dev.numerex.com/forums/viewforum/8/.

New FAST Release Coming Soon

We’ve just verified our latest build is ready for production. Revision 6885 to be exact. What does this mean? This means that next week we’ll be making a formal announcement about all the goodies in this latest build and how you go about getting it. Details to come, but stay tuned and we’re looking forward to the latest release!

Open Source Hardware for Wireless Sensor Networks = SquidBee

If you haven’t seen it all ready then check out SquidBee, which is an open source hardware initiative based on the ZigBee standard. We’re big fans of open source software and currently have several projects on GitHub. It’s exciting to see hardware vendors taking a similar approach. Another good example of this is BugLabs. While most of these platforms are being used by DIYers, it won’t be long before open source hardware becomes a more prominent player in commercial M2M deployments.

Twitter and M2M in the Home

I’m starting to see more and more uses of Twitter that bridge the gap between human and machine. More and more lifehackers are finding ways to make use of Twitter’s platform to send updates from their washing machine, power outlet, and even office chair (probably my favorite and most useless of the bunch). Machines should be able to tweet too, right?

http://mashable.com/2009/05/12/twitter-at-home/

Numerex on GitHub

For a majority of our development efforts centered around FAST (Foundation Application Software Technology) we’ve decided to make our code publicly available in GitHub. Here’s our Numerex profile page:

http://www.github.com/numerex

GitHub will be a great place for our customers to keep an eye on our Rails source code, see when commits are made, and new features are added.

Welcome to the Numerex Developer Blog

We’re excited to announce the official Numerex Developer blog. Our goal will be to keep you up to date with the latest developments covering all aspects of Numerex DNA. So tune in and subscribe to our RSS feed - once it becomes available : )