What’s new in Azure SQL Managed Instance at Ignite 2020

by Contributed | Sep 23, 2020 | Azure, Technology, Uncategorized

This article is contributed. See the original author and article here.

Azure SQL Managed Instance is a fully managed, secure, and always up-to-date SQL instance in the cloud, providing an ideal destination for modernizing your SQL Server applications at scale. SQL Managed Instance is part of the Azure SQL family of database services which include virtual machines and managed databases. Since its inception, Azure SQL Managed Instance (SQL MI) has been continuously improved based on customer feedback, aiming to meet critical requirements of organizations that migrate and modernize their applications in Azure.
In that spirit we are excited to announce several important capabilities coming to preview or general availability during Ignite 2020, which will bring significant improvements across multiple product aspects: performance, security and compliance, management experience, programmability surface area and application compatibility. This blog post gives you a quick summary of these new capabilities which will help you achieve more with Azure SQL Managed Instance.

Performance
General Availability: Major performance improvements for Azure SQL Managed Instances

We are thrilled to announce a set of major performance improvements for Azure SQL Managed Instances, which enable you to migrate your more performance-hungry database workloads to Azure SQL Managed Instance. These improvements include more than doubled transaction log write throughput for General Purpose and Business Critical instances and superior data/log IOPS for Business Critical instances. Increased performance predictability and stability for General Purpose service tier through better integration and enhanced tempdb performance are also included. These performance improvements are automatically enabled for all existing and future Azure SQL Managed Instances at no extra charge.

General Availability: Global virtual network peering support for Azure SQL Managed Instance

To provide you with an easier network configuration, all managed instances created in empty subnets will be hosted on virtual clusters that are enabled for access over global virtual network peering connections. This enables you to pair managed instances in a failover group configuration, in an easy and performant way, by simply connecting virtual networks in two different regions. By utilizing the global virtual network peering for your managed instances, you will save time through easy network configuration and offload your gateways from database replication traffic.

Compliance
General Availability: Configurable Backup Storage Redundancy option for data residency compliance

Now in general availability, locally-redundant and zone-redundant storage options have been added to Azure SQL Managed Instance, providing more flexibility and choice. These backup storage alternatives give cost-effective options for ensuring data is protected in case of planned and unplanned events.
In addition, if you need to meet specific data residency requirements and ensure that your data does not leave region of database deployment, this option will help you stay compliant.
Use Azure portal to configure backup storage redundancy easily or automate setting desired options in your PowerShell and Azure CLI scrips. For more information, see our Tech Community blog.

Easier management experience
General Availability: Enhanced management experience for Azure SQL Managed Instance

Azure SQL Managed Instance provide management operations that you can use to automatically deploy new managed instances, update instance properties, and delete instances when no longer needed. Most of the management operations in SQL Managed Instance are long-running but until now it was not possible for customers to get detailed information about operation status and progress in an easy and transparent way.
Through the introduction of a new CRUD API, the SQL Managed Instance resource is now visible from when the create request is submitted. In addition, the new OPERATIONS API adds the ability to monitor management operations, see operation steps, and take dependent actions based on operation progress.
Check out this blog post to learn how to effectively utilize new APIs in real-word scenarios.

Preview: Azure Active Directory (AAD) authentication configuration with non-admin privileges

Users can now assign groups to Directory Readers role in Azure Active Directory. This change will promote ease of use when setting up Active Directory Admin for Azure SQL Database and Managed Instance.

Effectively, AAD auth configuration for individual managed instances does not require Global Admin or Privileged Role admin’s action anymore.
With this change, setting AAD authentication managed instances will be much simpler and streamlined in large organizations.
At the high level, improved process will look like this:

1. Global Admin or Privileged Role Admin:
   1. Creates new Azure AD group where managed identities of Azure SQL Managed Instances will reside and assign “Directory Readers” role to this group. This action needs to be performed once per Azure AD tenant.
   2. Assigns group owners to the group – any Azure AD users without specific roles.
2.  Group owners then can add managed identity of an instance to the group.

Programmability and application compatibility
Preview: Machine Learning on Azure SQL Managed Instance 

Machine Learning Services with support for R and Python languages now include preview support on Azure SQL Managed Instance. When using Machine Learning Services in Azure SQL Managed Instance, you can run R and Python scripts to do data preparation and processing, train machine learning models in database, and deploy your models and scripts into production in stored procedures.

General Availability: Hosting catalog databases for all supported versions of SSRS in Azure SQL Managed Instance

Now in general availability, Azure SQL Managed Instance can host catalog databases for all supported versions of SQL Server Reporting Services (SSRS). Last year, SQL Server 2019 Reporting Services introduced native support for hosting catalog databases in SQL Managed Instance. Now you can also use SQL Managed Instance to host catalog databases for earlier supported versions of SSRS.

This is especially useful for fast migration of existing reporting solutions to SQL Managed Instance, without the need to test and adopt the latest version of SSRS. It also helps you get quick business benefits and then modernize further at your own pace.

To learn how to configure your SQL Managed Instance to host catalog databases for earlier supported versions of SSRS, visit the Tech Community blog.

Preview coming soon: Distributed database transactions spanning multiple Azure SQL Managed Instances

Distributed database transactions spanning multiple Azure SQL Managed Instances will be available in next few weeks. This will enable frictionless migration of existing applications and development of modern, multi-tenant applications relying on vertically or horizontally partitioned database architecture.

By utilizing distributed transactions, you will save time when migrating existing applications that require this capability as it eliminates the need to change application code and to perform extensive testing. If you develop new applications, you can benefit from partitioning data into multiple databases to overcome current sizing limitations of Azure SQL Managed Instance, while utilizing distributed transactions to keep partitioned data in a consistent state.

During preview, two methods of running distributed transactions will be supported using BEGIN DISTRIBUTED TRANSACTION statement from Transact-SQL code and using TransactionScope class from .NET code.
Check out Azure Updates to stay tuned with the upcoming public preview announcement!

Recently added features

We would like also to remind you about a couple of useful features that were recently added to Azure SQL Managed Instance. Visit the links below to learn how to utilize them in your application scenarios:

• User-initiated failover for application fault resiliency testing
• Backup storage cost saving options
• SCOM Management Pack for Azure SQL Managed Instance

Get started today!

• Get started quickly with Azure SQL Managed Instance using resources from our landing page on Azure.com.
• Find out more about how to migrate, modernize, and develop applications faster with Azure SQL by registering for our digital event Transform Your Applications with Azure SQL.
• Find out how SQL Managed Instance reduces management overhead with its virtually hands-free administration and gives you a low total cost of ownership in this Forrester Total Economic Impact Report.
• Read customer stories to learn how other customers utilize Azure SQL Managed Instance to improve their business 
• Try SQL Managed Instance today and let us know what you think! Feel free to submit your comments on this blog or to post product ideas and vote for existing suggestions at Managed Instance feedback page.

Deploy SAP IQ-NLS HA Solution using Azure NetApp Files on SUSE Linux Enterprise Server.

by Contributed | Sep 23, 2020 | Azure, Technology, Uncategorized

This article is contributed. See the original author and article here.

SAP IQ High-Availability can be achieved with IQ Multiplex architecture [ SAP Note 2477758], but not when deployed in NLS mode. As per the SAP First Guidance for SAP IQ 16.x, here, Multiplex Architecture is not available/evaluated for the SAP-NLS solution. This blog discusses one of the solutions to achieve SAP IQ-NLS Highly Available Solution using Azure NetApp Files on SLES 12.x.

Overview

High availability architecture allows an application to be resilient against hardware failure and to minimise downtime during maintenance activities. Many customers use SAP IQ to store historical data by extracting the old data from the SAP BW system to optimise the performance of the SAP BW system & achieve overall lower cost. It is significant to implement SAP IQ highly-available architecture to ensure data is online when required by the business processes. We have worked with our partners  [Accenture, Microsoft, SAP, SUSE] & customers to develop a highly available SAP IQ-NLS solution using Pacemaker along with other Clustering Components & Azure NetApp Files to eliminates a single point of failure. With this architecture servers [VM’s] are continuously monitored, and the workload is transferred to a healthy server when a fault or failure occurs using Azure Standard Load Balancer. The application is automatically restarted on a known healthy system to maintain business continuity and minimize unplanned downtime.

There is no official SAP IQ-NLS HA architecture reference from SAP or SUSE, and therefore, a custom solution must be created to achieve a solution to meet customer needs. The complete solution has been built and tested on the Microsoft Azure Cloud.

Architectural Overview

SAP IQ High-Availability Architecture on Azure VMs using Azure NetApp Files on SUSE Linux Enterprise Server.

Above SAP IQ-NLS Architecture contains:

• Two Azure VMs (mzapvldbiav01 & mzapvldbiav02)  running with SLES 12 SP4 for SAP.
• Three Azure VMs (mzapvlapsbd01, mzapvlapsbd02 & mzapvlapsbd03) running with SLES 12 SP4 for SAP with iSCSI Components which will be used as STONITH (SBD Devices) for Pacemaker Cluster Fencing.
• iav-ilb-sybase : Azure load balancer which will monitor health probe port and forward the traffic to the node where this health probe port is active.

The Cluster includes following resources:

• systemd:sapiq : Cluster resource for custom systemd service controlling sapiq database. Custom Systemd unit file which will take care of Start, Stop and Health check for Sybase IQ Database.
• azure_lb_health_probe : Cluster resources which create and listens custom port only on the active node.
• pri-ip_vip_63 : Cluster resource for floating IP which will be running only on the active node and works as virtual IP.
• anf_filesystem : Cluster resource containing shared azure netapp filesystem information to enable mounting filesystems only on the active node
• email-agent : Cluster resource which will trigger an email to recipients if any resource changes status like start, stop etc. 

Installation & Configuration Steps

1. Configuring iSCSI Target Servers
2. Create iSCSI devices on iSCSI target server
3. iSCSI Discovery on Cluster Nodes
4. Set up SBD Device on Cluster Nodes
5. Preparing operating system on cluster nodes
6. Cluster Installation
7. Joining Nodes to Cluster
8. Tuning Cluster Configuration
9. Preparing Custom Scripts for systemd service
10. ANF Shared Filesystem Configuration
11. Cluster Configuration

1. Configuring iSCSI Target Servers

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#set-up-iscsi-target-servers

The following Steps needs to be repeated on all iSCSI Target Servers

Step 1: Update SLES

Step 2: Remove packages

Step 3: Install iSCSI target packages

Step 4: Enable the iSCSI target service

2. Create iSCSI devices on iSCSI target server

 Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#create-iscsi-device-on-iscsi-target-server

The following Steps needs to be repeated on all iSCSI Target Servers

Step 1: Create the root folder for sbd devices

Step 2: Create the SBD device for Cluster Nodes on all iSCSI target servers

• Create block device

• Create iqn

• Adding iSCSI disks to the TPG

• Allow iSCSI initiators to connect to the iSCSI target.

 Step 3: Save the targetcli changes

 Step 4: Confirm the changes

 Step 5: Perform Steps 1 – 4 on all iSCSI Target Servers with their respective iqn.

3. iSCSI Discovery on Cluster Nodes

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#set-up-sbd-device

Step 1: Enable the iSCSI and SBD services on  Cluster Nodes.

Step 2: Change the initiator name on  Cluster Nodes

Initiator name should be unique on each server.

Step 3: Restart the iSCSI service on Cluster Nodes

Step 4: Connect the iSCSI devices

Step 5: Verify iSCSI Sessions

Step 6: Verify if disks are available on Cluster Nodes

Step 7: Retrieve the IDs of the iSCSI devices on Cluster Nodes

 4. Set up SBD Device on Cluster Nodes

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#set-up-sbd-device

Step 1: Create the SBD Device on Cluster Nodes

Step 2: Adapt the SBD Config on First Node in the Cluster

Populate the following parameters within the sbd file

Note: SBD_STARTMODE=”clean”, this will not allow any node to join to the cluster once node is fenced or rebooted. This is part of the Solution Requirement.

Step 3: Create the softdog configuration file 

Step 4: Now load the module

5. Preparing Operating System on cluster nodes

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#cluster-installation

Step 1: Perform system update on all Cluster Nodes

Step 2: Install components needed for Cluster resources on all Cluster Nodes

Step 3: Configure the operating system on all Cluster Nodes

Increase the maximum allowed processes

Edit the configuration file

Change the DefaultTasksMax

                                                      to

Activate this setting

Test if the change was successful

 Step 4: Reduce the size of the dirty cache on all Cluster Nodes

Change/set the following settings

Step 5: Configure NTP on all cluster nodes and confirm

stop ntpd daemon

edit the /etc/ntp.conf to make appropriate entries of NTP servers.

Now  start the ntp daemon.

check the NTP status

 6. Cluster Installation

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#cluster-installation

Step 1: Install Cluster on first node

Step 2: Respond to the following on-screen instructions

7. Joining Nodes to Cluster

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#cluster-installation

Step 1: Add Secondary node to cluster

Step 2: Respond to the following on-screen instructions

8. Tuning Cluster Configuration

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/high-availability-guide-suse-pacemaker#cluster-installation

Step 1: Change hacluster password on all cluster nodes

Step 2: Adjust corosync settings on First Node

Change token to 30000 and consensus to 36000

Restart the corosync service

9. Preparing Custom Scripts for systemd service

Step 1: Create Systemd Service which can initiate start and stop scripts of the SAP IQ DB and place it under /etc/systemd/system/

sapiq.service should contain the following

When Start initiated on sapiq.service, the following script will be called

When Stop initiated on sapiq.service, the following script will be called

Step 2: Creating sapiq_start.sh & sapiq_stop.sh

sapiq_start.sh:

Note:  Contact your NLS Admin for DSN Value (SAPIQDB_IAV, in this case) and replace the same in the above script.

sapiq_stop.sh:

The above two scripts should  be placed under /usr/bin

10. ANF Shared Filesystem Configuration

Reference Link: https://docs.microsoft.com/en-us/azure/virtual-machines/workloads/sap/sap-hana-scale-out-standby-netapp-files-suse#operating-system-configuration-and-preparation

Step 1: Prepare the OS for running SAP IQ NLS  on NetApp Systems with NFS

Create configuration file /etc/sysctl.d/netapp-hana.conf for the NetApp configuration settings.

Within the file, keep the below entries

Step 2: Create configuration file /etc/sysctl.d/ms-az.conf with Microsoft for Azure configuration settings.

Within the file, keep the below entries

Step 3: Adjust the sunrpc settings

Within the file, keep the below entries

Step 4: Ensure to set the NFS domain in /etc/idmapd.conf on the VM to match the default domain configuration on Azure NetApp Files: defaultv4iddomain.com.

11. Cluster Configuration

Following cluster resources needs to be configured

• Floating IP: Configure Virtual IP which is always available on active node. Here we need to use frontend IP address of Azure Load balancer as Floating IP in Cluster.

• Health Probe Port: Configure 61000 as health probe port  which is always available on active node. Azure Load Balancer always monitor this port and sends the traffic to the node where this port is running.

• Systemd Service: Cluster Primitive which controls start & stop operations of SAP IQ Database.

• ANF Shared Filesystems: Individual Cluster Primitive for each ANF Shared filesystem for log, system, dbspace & database. These four filesystems will be shared between two cluster nodes.

• Email Alerting: One of the easiest ways to get notifications of any cluster resource events  is to add a ClusterMon resource. Whenever any resource is started or stopped, the ClusterMon resource will send notification to the custom script which will trigger email to configured recipients.

And clone the resource “email-agent”

Post-Configurations

Create the custom script and place under /usr/bin

Within the file, place the following content

• Grouping of Resources: Grouping of resources will make all child resources to start or stop together on same node.

Grouping resources which includes ANF Filesystem, Floating IP, Health Probe Port & Systemd service. In this case, group name is  “grp_anf-sapiq_systemd-azure_lb-vip_6” with migration threshold set to 1**.

**  Migration threshold will allow resource  to restart on failure with the provided value on the same node.

• Order Constraint: As the group containing systemd resource rely on shared file system to start the database, file system group should start before the group with systemd resource then health probe and virtual IP must start.

• Resource Stickiness: This is to prevent healthy resources from being moved around the cluster.

 Note: Defaults don’t apply to resources which overwrite them with their own defined resource stickiness value.   

Complete CRM Configuration as follows 

primitive anf_filesystem Filesystem

        params device=”170.162.118.4:/Database” directory=”/usr/sap/IAV” fstype=nfs options=”rw,vers=4,minorversion=1,hard,sync,timeo=600,rsize=65536,wsize=65536,intr,noatime,lock,_netdev,sec=sys”

        op stop timeout=60s interval=0

        op start timeout=60s interval=0

        op monitor interval=20s timeout=40s

mzapvldbiav01:~ # crm configure show pri-ip_vip_63

primitive pri-ip_vip_63 IPaddr2

        params ip=170.162.92.63 cidr_netmask=24 nic=eth0

        op monitor interval=0

mzapvldbiav01:~ # crm configure show grp_anf-sapiq_systemd-azure_lb-vip_63

group grp_anf-sapiq_systemd-azure_lb-vip_63 anf_filesystem pri-sapiq_systemd azure_lb_health_probe pri-ip_vip_63

        meta target-role=Started migration-threshold=2

mzapvldbiav01:~ # crm configure show

node 1: mzapvldbiav02

node 2: mzapvldbiav01

primitive anf_filesystem Filesystem

        params device=”170.162.118.4:/Database” directory=”/usr/sap/IAV” fstype=nfs options=”rw,vers=4,minorversion=1,hard,sync,timeo=600,rsize=65536,wsize=65536,intr,noatime,lock,_netdev,sec=sys”

        op stop timeout=60s interval=0

        op start timeout=60s interval=0

        op monitor interval=20s timeout=40s

primitive azure_lb_health_probe azure-lb

        params port=61000

primitive email-agent ClusterMon

        params extra_options=”-E /usr/bin/crm_email.sh -e ramesh.petla@accenture.com”

        meta target-role=Started

primitive pri-ip_vip_63 IPaddr2

        params ip=170.162.92.63 cidr_netmask=24 nic=eth0

        op monitor interval=0

primitive pri-sapiq_systemd systemd:sapiq

        op start timeout=60 interval=0

        op stop timeout=60 interval=0

        op monitor timeout=100 interval=10

primitive stonith-sbd stonith:external/sbd

        params pcmk_delay_max=30s

        meta target-role=Started

        op start interval=0

group grp_anf-sapiq_systemd-azure_lb-vip_63 anf_filesystem pri-sapiq_systemd azure_lb_health_probe pri-ip_vip_63

        meta target-role=Started migration-threshold=2

clone cln_email-agent email-agent

order ord_anf-sapiq_systemd-azure_lb-vip_63 inf: anf_filesystem pri-sapiq_systemd azure_lb_health_probe pri-ip_vip_63

property cib-bootstrap-options:

        have-watchdog=true

        dc-version=”1.1.19+20181105.ccd6b5b10-3.16.1-1.1.19+20181105.ccd6b5b10″

        cluster-infrastructure=corosync

        cluster-name=cls_iqdb_iav

        stonith-enabled=true

        last-lrm-refresh=1594989262

        maintenance-mode=false

rsc_defaults rsc-options:

        resource-stickiness=1000

op_defaults op-options:

        timeout=600

        record-pending=true

Monitoring  and Managing the Cluster

CRM Status: The status of the cluster can be checked by issuing command “crm status”

CRM Monitoring: Real time changes in the cluster can be monitored continuously  by issuing command “crm_mon” on any node which is part of cluster.

The above similar things can be monitored using HAWK Web Console.

To access HAWK Console, enter the following URL with username “hacluster”

where 170.162.92.63 is the ip address of any cluster node or virtual IP. 

Resources Status:

Node Status:

Node Cleanup Script: Whenever nodes reboots because of Kernel Panic or any other issues, run the crm_node_cleanup.sh script which is placed under /usr/bin  to clean the node so that it can join to Cluster.

For mzapvldbiav01:

For mzapvldbiav02:

The above scripts are made only for saving time, otherwise, one can run individual commands to clear node messages.

Testing Scenarios

Scenario 1: Stop database on mzapvldbiav01

Stop the IQ DB on database mzapvldbiav01 gracefully. This should failover the database into mzapvldbiav02 and promote it as a master.

The cluster detects the stopped primary IQ database (on mzapvldbiav01) and marks the resource failed.  The cluster promotes the secondary IQ NLS database (on mzapvldbiav02) to take over as primary.  The cluster migrates the IP address to the new primary (on mzapvldbiav02). The cluster “failed actions” are cleaned up after following the recovery procedure.

Before Test: Cluster Status

Stopping Database:

After Failover:

Resource Cleanup: As the resource pri-sapiq_systemd is failed to restart on mzapvldbiav01, need to perform resource cleanup on mzapvldbiav01 for  pri-sapiq_systemd. So next time if mzapvldbiav02 is down for some reason, resource can migrate to mzapvldbiav01 without any issues.

Cockpit View:

Scenario 2: Crash mzapvldbiav01

Simulate a site crash situation using kernel panic (echo ‘b’ > /proc/sysrq-trigger) on database  server mzapvldbiav01. This should failover the cluster resources to mzapvldbiav02.

The cluster detects the failure on mzapvldbiav01 and declares it UNCLEAN. The cluster fences mzapvldbiav01. The cluster declares the failed node mzapvldbiav01 as OFFLINE. The cluster promotes the secondary IQ database  on mzapvldbiav02 as active.

Before Test: Cluster Status

Simulate Kernel Panic on mzapvldbiav01:

After Failover:

As SBD fencing is used, we need to clear reset message on mzapvldbiav01 once root cause for kernel panic is identified and fixed. The reset message can be cleared with the help of the script /usr/bin/crm_node_cleanup.sh on respective nodes.

After Cleanup:

The node mzapvldbiav01 status will change to Online from Offline.

Cockpit View:

Troubleshooting

Resource Fail count Cleanup: Resource needs to be cleaned if any resource has failed errors like follows

In the above screenshot, we have errors associated with systemd resource “pri-sapiq_systemd”

Clearing Temporary Constraints: When we move any resource to other node manually, then location constraints will be created. These needs to be cleared if we need cluster to handle resource movement operations automatically.

Log Files:

Pacemaker Log File:

The systemd journal can be used to retrieve Pacemaker logs:

Alternatively, /var/log/messages can be referred for other errors

Decision Tree