Amazon EC2 vs Microsoft Azure, Part III

 

In the past 14 days, the migration of technology was amazingly easy (okay, maybe for the novice, but for me, it was just a matter of patience and timing).

 

Based on ones experience it was great to work with other professionals who did not have all the experience but enjoyed the ride.

 

What did we learn in the move?

 

1. Working with stakeholders was fun. Knowing they trust you to do what is right, necessary, done by a master of all things technology, a great tester, and making everything painless, and the results would be a profit making opportunity and an immediate cost savings was enjoyable.

2. Migrating applications from SQL OLE-DB 2003/2007 to SQL 2010/2016 libraries was fun.

3. Migrating applications from old mail handlers to new mail handlers was fun.

4. Migrating email from 1998 processing engines to 2016 fully encrypted and fully compliant email servers was fun! The DNS and all the Email checkpoint requirements was fun to figure out. Make sure your A record forward for your email server is set before you ask for a Reverse DNS entry. They require the forward to be in place first though they should not because it causes a longer than required blackout period for a live email server migration.

5. Migrating phone lines and services was fun.

6. Migrating SSL technologies, certificates and adding security layers to E-Mail systems and connectors was fun.

7. Backing up old servers and only migrating data that was needed and leaving old stale data in backups and not migrating unnecessary data was a great feeling.

8. DNS is fast to update these days. In the past the TTL for DNS records was two weeks. Now that it is 5 to 30 minutes, the black out and dark time period is much less.

9. Turning down three (3) servers that have been running for 10 years was fun! Sad to see them turned off and yet I am ready bring them back online in a different configuration and repurposed.

10. Between security and performance, and unknown expectations in service providers service level agreements that is something of a great expectation.

11. Setting up, testing and verifying backups and snapshots was fun.

12. Billing snapshots and estimations were cool. Billing and aggregate activity is easier to see in AWS than Azure. Azure has a terrible methodology in billing for services.

13. Network bandwidth allowed boosts for uploads and installs was tolerable.

14. System Event Logs (application, security and system) have stayed cleaned and perfect.

15. Firewall and application protection filtering was easy to manage and confirm and test.

16. Overall, the final move saved our company some 75% in the bottom line, and the 25% that had no savings was still passed along to our clients and their needs. Basically, we concentrated and reduced total cost of ownership (TCO) - a very overused term in the past decade.

17. Only two clients had some timing issues which we were able to navigate around.

18. Document your steps in advance as the plan and then follow it, and then adjust it as you go.

19. Update your recurring billing items to reflect the changes.

20. Make sure you time your migration of each server and application when the users are definitely offline or mostly not needing services, even if it takes you from 11pm to 2am to do it. Just be awake and ready by 7am to see if anyone has any issues one cannot determine regardless of remote testing from various locations across the globe.

21. No matter, have fun doing this migration. Not everyone can do it well and with ease. Just set expectations properly and pull the trigger and make sure you aim perfectly.

22. Oh, don't forget to set your new reminders for new changes on services through other vendors. I accidentally forgot to renew an SSL certificate that actually expired the day after all the migrations, and while the outage was only a few hours, the impact was 350 clients who were asleep, and 2 administrators who were just waking up when I noticed the SSL expiration notice. the Good thing the certificate was already renewed and waiting to be downloaded and applied to the servers. 15 minutes of hurry up and voila, back in business. I just notified the administrators that were 2 and 3 hours ahead of me in time zone who might have been impacted and did not say anything.

23. Make sure your new backups are in place and business continuity and disaster recovery methods are in place. I had to make an adjustment to two (2) servers that we partially migrated, rather than moving to the cloud we just moved to another network that was serving and IoT array so we stayed on premises for the moment like one last co-located server.

more to come...

 

SIDE NOTES

 

For me, it was just another walk in the park. For 30+ years, migrating systems and upgrading is not for the faint of heart. It requires research, planning, testing, and the courage to pull the trigger and just do it.

 

What will become of Integra Telecom? After 41 applications and 7 servers moved in the last 14 days from the ole Integra Telecom network to the Amazon and Charter Network, who else needs to switch? Integra switched out to be AllStream in the last 12-24 months, and now that after my 9 years with Integra, I am happy to say, the quality and price of their service was spectacular during those times.

 

The long road of networks and migrations experienced: Prior to Integra from 2009 to 2018, we had Great Basin Internet Services from 1998 to 2009. Somewhere in between 1995 til now and along the way we experienced opportunities with The-OnRamp with Rich Brown, DesertLinc with Cliff Hayden, Don from Cyber Highway, Aztech Cyberspace with Josh and Nick, Brooks Fiber with Helen, ATG with Bruce and Kim, Pyramid.net with Bob and Todd, Great Basin Internet Services with Bruce, Redundant Networks with Janice and KC, Clearwire with Sylvia, Level 3, Highwinds, IO, Comcast, Highlands Wireless with John, Sky Fiber, AT&T Gigabit, Utility Telecom, Charter Spectrum Cable, and Josh with Charter Fiber, etc. If you really want to know, ironically, and historically, we started with Virtual Cafe by Peter Dewey in Snohomish Washington in 1995.

 

Our move continued our solutions between the remote data center and on site data center. This was my version of a hybrid solution 20 years ago, long before it was cachet in the past 10 years.  On premises and in the cloud, is the way to go. Keep your intellectual proprietary solutions local and hard-walled and your non-core yet business essential things in the cloud where bandwidth is not a premium.

 

more to come...

 

Would Microsoft Azure be just as easy. Not Really. #4 would have been a pain, and

in case you want to read what they have to say which is about what I have said:

 

https://aws.amazon.com/cloud-migration/