Azure IaaS VMs and the Three (now Four!) R’s

When you have a problematic IaaS VM (won’t start, won’t stop, can’t RDP even though it worked just yesterday….) and you’ve exhausted your usual tricks, turn to the Three Four R’s (the “did you reboot?” of Azure).

  1. Restart – Most users will think of this, just be sure you restart (and I mean a stop and start, not the restart button) from Azure (portal or PS) and not from the VM.  The Azure restart will give the fabric a chance to look for issues and self-heal.  Note: If you have a classic VMs (old portal) this also applies to your cloud service.  I’ve seen VMs acting up due to issues with their cloud service.  You can try restarting the cloud service….but keep in mind all hosted VMs by the cloud service will be restarted as well!
  2. Resize – Resizing (esp. if you increase the VM size to the largest possible), it will recreate certain elements of the VM <-> Fabric connection and could even move you to a new cluster node on the backend.   Resize, test to confirm it’s working, and then size back to your original size.  You will get charged at the higher VM size rate, but if it’s only for 15min that’s a minimal cost.  *Note: if you’re in ARM (VMs deployed via the new portal, not a classic VM) you can directly redeploy to a new cluster node using Azure powershell: (step 4 below)
  3. Recreate – This one sounds scary, but if you make a note of your current configuration FIRST the recreation is quick (generally <20min) and relatively painless.  If you need to move your VM to a new cloud service, VNet, etc. or are having those “it’s just acting up” issues this is a good troubleshooting step to try out.   You basically are removing the Azure components and then recreating the Azure components (modifying if needed) — all while leaving your disks untouched.  See  for step-by-step instructions.  *Note: The link is specific to ASM (classic portal) but the premise works for both classic and ARM VMs.
  4. Redeploy – Available in the new portal only for ARM VMs (not classic).  Effectively the same as a resize, only guaranteed that you change nodes.

If these all fail, and you already confirmed there are no outages that could impact you (, it may be time to engage Microsoft.

See also

*previously posted at

Azure VMs need Internet Access

When customers move into the cloud, they tend to mimic their setup on-prem.  Not a bad thing, but when it comes to blocking internet access for servers this can create some unusual problems.

If you are using network security groups (NSGs), user defined routing (UDR), or forced-tunneling be sure to put in an exception for your Azure data center IP ranges, as lack of connectivity will impact many services including these:

  1. VM Extensions see
  2. Azure Backup see
  3. Monitoring Agent/Extension see
  4. KMS

Update 16 Aug 2018 – The use of service endpoints will limit the damage of blocking internet access.  Ensure all services you use/require are covered by service endpoints before blocking internet access.

*previously posted at

Connecting to MySQL for Azure Site Recovery (ASR)

When using ASR to replication VMware or physical machines into Azure two roles are required – the configuration and process servers (often combined on a single server) – to help coordinate and facilitate the data replication (  On the configuration server, configuration data is stored in a MySQL database.  *at this time this is a requirement to use MySQL, other databases types are not supported.

There are several scenarios when you may need to verify or modify data stored in this database.  Below are samples for your reference.

Note – database modifications will impact ASR and should be done with care

Login to MySQL and Connect to the ASR Database

from a command prompt:

mysql –u root –p  (you will be prompted to enter the password specified during installation)

show databases;  (will list all databases for your reference)

use svsdb1; (selects the ASR database so future queries will run against it)



To list all machines registered with the configuration server (CS)


select id as hostid, name, ipaddress, ostype as operatingsystem, from_unixtime(lasthostupdatetime) as heartbeat from hosts where name!=’InMageProfiler’\G;



To Cleanup Duplicate/Stale Entries



To Update the IP of a Machine

update hosts set ipaddress='[new address]’ where ipaddress='[old address]’;

example, update hosts set ipaddress=’′ where ipaddress=’′;


*previously posted at

Azure IP Ranges

*Updated June 15, 2018*

For a myriad of reasons it’s nice to know what IPs you can expect to see coming to/from your Azure space.  Below is a quick cheat sheet.

Microsoft Azure Datacenter IP Ranges

updated 20 Aug 2018, thanks to Michael Ketchum of Microsoft for the additional information

The XML file now breaks down the IP ranges as follows:

  • “<SERVICE>” = Includes all IP’s for that service across all regions in the applicable cloud
  •  “<SERVICE>.<REGION>” = Includes all IP’s for that service in a specific region
  •  “AzureCloud.<REGION>” = Includes all IP’s/and services for that region
  •  “AzureCloud” = Includes all IP’s/and services for that cloud, such as Gov, commercial, etc.

The “Secret Azure IPs” you MUST Include – and

Office 365 URLs and IP address ranges

Office 365 US Government: Endpoints for US Federal and US Defense Clouds (preview)


*previously posted at

Using OMS to Alert on Azure Service Outages

If you are unable to alert from the Azure Portal, or simple wish to have all your alerting from one source, consider leveraging OMS (Operations Management Suite).  With the free tier option (7 days of day retained) there is no additional cost!

Azure Service events are logged automatically in the Azure Portal –> Monitoring –> Activity Log (only incidents believed to impact your subscription(s) will be listed).  This article will show you how to use OMS to review and alert on these events.



Setup OMS (if you do not already have an OMS workspace)

1. Create a new workspace –

  • select the free pricing tier unless you have further plans for OMS


Configure OMS to Pull the Azure Activity Logs

1. Add the Activity Logs Analytics solution – (only steps 1-4 are required)


Setup Alerting

1. Open the OMS portal (URL varies based on your cloud)

2. Click on Log Search

3. In the query window, enter: Type=AzureActivity Category=ServiceHealth

    • This will looks for alerts from the Azure Activity logs of type Service Health.  This is how Azure Service outages are categorized in the Azure Activity Logs
    • it is OK if no results are returned.  That just means there were no Azure Service Incidents that impacted your subscription the time range.


4. Click Alert in the top left

5.Configure the alerting options (see for more details)


*The alert looks every 15min (alert frequency) for events matching the query that were raised in the past 15min (time window).  If there are more than 0 found (number of results), then an email is sent to all recipients listed.  These emails do NOT need to be associated with an Azure logon, etc.  Any publically routable email address will work.

Your recipients will now receive an email for Azure Service incidents.  It will look something like this:


*previously posted at

Using Azure CLI 1.0 to Copy a File between Subscriptions

note: examples are from the Azure Government cloud but the command used will work in all clouds


Goal: Use Azure CLI 1.0 to copy a blob file between to different subscriptions

Syntax:  azure storage blob copy start “sourceSAS” destContainer -a destStorageAccount -k destStorageKey

Example: azure storage blob copy start “” vhds –a secondstorage -k zcSRtkJO9LzJuiYbOgoRW6Fgr3lS7lIFIEvIb3hbzJ62XBmZl5Igg1zfogNee8FtwGNGoJ6ADr7kAls6b+wJNQ==

note: the SAS and storage account key are used for access to the storage accounts, subscription access is not required to execute this command.


Now let’s break it down….here’s how you gather each of the required inputs.

1.  SourceSAS – the SAS is the source file URL + the Shared Access Signature (SAS).  For example if your source URL is and your SAS Token is ?sv=2016-05-31&ss=bfqt&srt=sco&sp=rwdlacup&se=2017-05-16T06:59:23Z&st=2017-05-15T22:59:23Z&spr=https&sig=lyzka%2F3ID1qFhLeoxlDcjBpNuHDB001qWL0ubiD66wo%3D your SourceSAS is  Note: There are no spaces when you join the two items and you’ll need to put the string in quotes when used in AzureCLI.

  • Source URL – there are a few ways to get this, but the simplest is via the portal.  Browse to your storage account –> blob –> your container –> your file.  A new blade opens and the URL is the second item listed.


  • SAS – again the simplest way to generate the SAS token is via the portal.  Browse to your storage account –> Shared Access Signature, update the values (the default will work, but it’s more secure to restrict the SAS Token to only the time frame and resources needed), and then click “Generate SAS”


2.  Destination container – this is the name of the container only (not a URL) that already exists in the destination storage account

3.  Destination Storage Account – this is the name of the storage account only (not a URL) that already exists in the destination subscription

4.  Destination Storage Access Key – there are a few ways to get this, but the simplest is via the portal.  Browse to your storage account –>Access Keys and copy either key1 or key2.

…and a special thanks to Madan Nagarajan for his sourceSAS breakthrough!


*previously posted at

Working with Azure ARM VMs, Images, and Unmanaged Disk (Storage Accounts)

A lot of the pre-managed disk documentation has become hard to find.  Below is a cheat-sheet on where to find the documents you need to work with Azure ARM VMs, images, and storage accounts.

Create an Azure VM from custom image (VHD) –

Create an Azure VM from existing VHD –

Create an Azure VM Image (and VM) from existing Azure VM –

*previously posted at

Moving files between Azure Storage and RHEL

There are several options when you want to move files in or out of an Azure Storage account to a Red Hat Linux (RHEL) server.  Below is a quick break down of the most commonly used options.

Azure CLI

Azure CLI is design to run on Linux or Windows so is the ideal tool when a Linux machine is involved.  Azure CLI 2.0 ( is the latest and preferred version.  See for syntax.


#Start the copy of the blob

az storage blob copy start –account-name ${storage_account} –account-key ${key} –source-account-name ${source_account} –source-account-key ${source_key} –source-container ${blob_container} –source-blob ${blob_name} –destination-container $target_container –destination-blob $target_blob_name

#Now wait while this copy happening. This could take a while.

while [ $(az storage blob show –account-name ${storage_account} –account-key ${key} –container-name $target_container –name $target_blob_name| jq .properties.copy.status| tr -d ‘”‘) == “pending” ]; do

progress=$(az storage blob show –account-name ${storage_account} –account-key ${key} –container-name ${target_container} –name ${target_blob_name} | jq -r .properties.copy.progress)

done_count=$(echo $progress | sed ‘s,\([0-9]*\)/\([0-9]*\),\1,g’)

total_count=$(echo $progress | sed ‘s,\([0-9]*\)/\([0-9]*\),\2,g’)

progress_percent=$((100 * $done_count / $total_count))

echo “Copying: $progress ($progress_percent %)” && sleep 5


echo “Copying done”


AzCopy on Linux is a command-line utility designed for copying data to/from Azure Blob and File storage using simple commands.  See  Specific steps for the file copy are here.

To Install on RHEL, you first need to install .NET Core 1.1.1

  1. sudo yum install rh-dotnetcore11
  2. scl enable rh-dotnetcore11 bash
  3. source scl_source enable rh-dotnetcore11

Then you install AzCopy per the article.

  1. wget -O azcopy.tar.gz
  2. tar -xf azcopy.tar.gz
  3. sudo ./

Example from RHEL 7.4:



The final approach is to downloading is to generate a SAS token, append it to the blob URL and simply use wget in Linux.

See for details on how to generate the SAS token and append it to the the blob URL (aka, “sourceSAS”).


*previously posed on

Azure Government – Missing Features or Services?

If you’ve experimented with the general Azure cloud (often called Microsoft Azure Commercial or MAC), it can be a shock when you move into Microsoft Azure Government (MAG) and notice not everything is the same.  Today I’m briefly going to discuss why there are feature differences, how to track them down, and how to lobby for the features you need.

Note: in this blog I use the terms service, offering, and feature interchangeably.  There is debate (even internally!) on the difference between the terms, but my goal here is to cover both new offerings and extensions to existing offerings.

Is the feature available in MAG at all?

Many features are turned on by region (a physical data center).  The first thing is to check and see if the offering (say a DS series VM) is available anywhere in MAG.

  1. Goto
  2. Select only the Azure Government region checkbox
  3. You can now see this VM is only available in the Virginia data center.


If possible, deploy in that region so you can use the feature/service.  However, we understand that is not always possible.  Before any new large-scale deployments you should take a look at the list of services per region and ensure you use the region that best offers the services you need.  You don’t want to end up putting yourself in a situation where you have to wait on required services to be deployed by Microsoft!

To stay current on new releases (and they happen at least once a week!) I recommend subscribing to the Azure Government blog at 

The features/service is not available where I need it!

Why not?  This is the big question I get from customers and of course there is no one easy answer.  These are the main reasons cited:

  • Features often depend on physical hardware and so they often deploy to each region at slightly different times depending on the hardware deployment.  I.e. with a new VM size/series
  • Some features are delayed as we work on the required compliances or attestations.
  • Engineering wants to release features as quickly as possible.  If the product can be deployed/engineered the same in MAC and MAG they are often released at the same time.  However if significant changes are required for MAG they don’t want to delay the MAC release.  In those cases the feature most likely will be deployed to MAC and then MAG after the required modifications are made.  Often features are delayed 6-9 months in MAG.
  • There is no/limited demand expected in MAG.

But I need it!

First, contact Microsoft.  You can talk with your sales team, account team, or even open a case via the portal (  We may have an estimate that can be shared if you are under NDA.

Second, vote it up!  At you can nominate new features, add your vote to requests made by others, and generally make your voice heard.  This is a VERY powerful tool, even though it looks innocuous.  Cloud computing is about consumption and by offering the services you want, we increase consumption.  However knowing exactly what customers want is complicated and can come down to a “gut feel.”  By assigning metrics (used for good this time!) we can see exactly how important a feature or change is to the community and react accordingly.  Azure Engineering does review this and uses this data in their planning.  While a highly “voted” item isn’t guaranteed to be implemented quickly, it will increase the visibility and urgency around it.

Here’s a few tips for submitting/voting on improvements.

  1. Include hard facts.  We want numbers, timelines, and project implications.  I.e. if the M series were available in Texas we would deploy an additional 10 VMs by the end of 1Q 2018.  This gives Microsoft an idea of the urgency/timeline, infrastructure required to support the request, and the impact to our accounts.
  2. Include clear details.  I.e. if this is available in MAC, include the blog announcement or documentation too so we know exactly what you are requesting.  Often features have similar names, vary between cloud providers, etc. resulting in confusion.
  3. Include the consequences.  If this request isn’t met, what is the result?  I.e. if Service Map is not made available in MAG OMS will we not migrate our monitoring of 8000 servers into OMS.
  4. Share with your friends!  Votes count, so if you add a request have all your peers vote for it too.
  5. And be realistic.  Asking for a feature not even planned for MAC is unlikely to be approved for MAG.  Asking for a feature fulfilling a clear business need is much more likely to get backing.

If you get stuck or have questions, feel free to comment!

*previously posted on