The third day at Ignite was kind a hard to start up, it were long day’s and fun long nights but 2 double espresso kind a pushed me out of my morning dip. Ready to start the day!
Azure High Performance Networking:
This sessions was initially not about new stuff. It’s was more to make things more clear about Azure networking. Near the end there was a lot of new stuff about ExpressRoute though!
Public and Microsoft Peering
Earlier I hear some noise from several people that the Office 365 peering, or Public Peering was to be canceled. But now we know that it’s not cancelled but that the 2 peerings have merged. That makes things simpler, but also more complex, because one of the most issue’s I hear customers talking about is that they don’t want to peer with all Azure or Office 365 services and now there is no choice in those either. It’s either none ore all in! But Microsoft must have heard this complaint because the came up with a new feature for ExpressRoute called Route Filters. With the filters you can choose what routes you want advertise to use only the service you want over the ExpressRoute connection. Nicely done! 🙂
Finally monitoring on ExpressRoute!
Monitoring is the next thing that is announced for ExpressRoute. Most of the devices are not managed by companies and not all ISP’s offer monitoring services. Now with the new ExpressRoute Network Performance Monitor (NPM) you can get a lot more insight in your ExpressRoute connection. You can monitor latency and throughput now within the Azure portal. They also included a Topology Dashboard to view your topology from onprem to azure and al hops it’s taking. You can do some tests to from certain endpoints. Really making progress here!
But there is more 🙂 a new view is created to view some circuit history to view the stability of your connection. If you want al this stuff you have to take an OMS license, that is the only part I am disappointed about. It should have been included in the ExpressRoute package if you ask me.
Azure Stack Security and Compliancy
Ever since it got launched I wanted to figure out how it all works together with Azure Stack turns out it was a wasted of time according to Jeffrey Snover 🙂 I quote: “Internals are internals! We will change a lot of stuff in the future and we aren’t going to tell or document the internals”.. oké understood. As an Ops guy I don’t like it, but hey.., its a new age..
The sessions had some overlap with a session from the day before, it was more about the compliancy and why they closed it down as much as possible. Overall most of the Azure Stack session had a lot of redundant information in it. It think they also tried to spread the load a bit but most of the sessions ware not that crowded as it is still really new to a lot of people.
Azure Stack and DR
So it you have you goodies delivered by either HPE, Dell, Lenovo or Cisco and are fully in production things could go ugly. As for the Azure Stack it self there are some DR procedures you have to take so you are ready to get it up and running again when shit hits the fan!
Lets start with what the Azure Stack DR backup is not doing. That is making a backup of your IaaS or PaaS data you are in charge to arrange something for that. Also switch config and OEM data. That is the responsibility of the OEM. It does give you the ability to create backup sets that are encrypted because it contains sensitive data. You have to enable the DR Service through powershell. The data is placed on an external file share. It are all full backups with data sets of approx. 10GB. You need about 1TB of storage for one Azure Stack system and data is saved for 7 days as recommendation. After 7 days it’s a manual deletion.
Once backup is configured you can view the status in the portal under the Backup Resource provider or in the File Share.
There are several vendors how will over IaaS and/or PaaS protection like Veeam, Comvault, Carbonite and more. Beware, it’s al very new and building in progress.
That’s it for Day 3 then!
Cheers!
Pascal Slijkerman