Each distributed system platform has unique nuances. In Part 1 of this blog, I demonstrated how having a single view to manage your multi-vendor SAN infrastructure helped ensure performance and understand the overall health, performance and capacity. What is equally important to these common views is a solution that is capable of getting the detailed performance data capable of supporting vendor specific architectures.
New storage system platforms are popping up every year, and it’s impossible to stay ahead of all of them and provide the detailed, intelligent, performance views necessary to manage your SAN infrastructure and prevent incidents. However, IntelliMagic Vision supports a wide variety of SAN platforms for which we provide our end-to-end capabilities.
EMC VMAX Auto-Tier Breakdowns
For EMC, there are several aspects that are important in terms of performance management:
1. Visibility into the Storage Group Performance and their associated Tiers (Policies)
2. Ability to drill down from Storage Groups to their volumes and see volume tier breakdowns
3. Ability to see front-end imbalances and quickly identify associated hosts/storage groups/volumes
IntelliMagic Vision makes all 3 of these activities easy.
HDS Detailed Performance Statistics
Getting detailed performance statistics out of HDS systems is not easy. The SMI-S and RESTFUL APIs do not provide enough detail, so we use the native CLI to get detailed performance data and it allows us to provide a rich set of performance and configuration information for HDS and HP Enterprise systems.
In these charts, we show how you can drill down from an HDS storage system to its storage pools and then to its volumes to see which volumes are driving load on a pool and impacting the performance of a storage system. This is simply a matter of drilling down from the top layer of the storage system to the volume level.
IntelliMagic Vision provides continuous detailed statistics collection that supports rich volume statistics, storage pool aggregations from volume data, host views of performance, detailed port performance metrics, and RAID group utilization.
HPE 3PAR AO Policy and CPGs
When managing the performance of HPE 3PAR environments there are a couple of challenges:
- Understanding the Adaptive Optimization Policies and how they affect workload performance
- Understanding the capacity reporting as 3PAR is different than any other enterprise array in the way it uses and reports capacity.
In the first chart, we are looking at the I/O rate for a particular AO policy’s volumes. This provides us the information we need to understand the performance of a particular AO Policy.
From the volume, we can drill down to see additional information regarding the Common Provisioning Groupt (CPG) breakdown which we show in the second chart.
In this case, all of the data resides on a single CPG, CPG_900gb_R6_6plus2. This CPG consumes data from the 450 GB and 900 GB drives. If they would like to improve the performance of this system then they will need to place this volume in a policy that contains SSD/Flash devices.
IBM Spectrum Virtualize: Mdisk to Back-end Volume Mapping
In this next view, we will drill down from the IBM SVC Storage Pool to its external Mdisks, to the associated back-end volumes. In doing this, we switch from the SVC perspective to the front-end of the back-end storage system seamlessly.
In the first view, we see the external Read I/os per second to the mdisk104.
In the second chart we drill down to the volume associated with the mdisk, vol-0000000000109.
As you can see, the number of read I/Os matches between the external reads on the mdisk and the front-end (different data source) reads on the back-end storage controller. This type of visibility allows you to understand if there are any issues on the underlying back-end volumes and provides complete end-to-end visibility of the Spectrum Virtualize stack.
NetApp: Physical and Logical Awareness
When managing NetApp performance, it is important to understand both the Physical and Logical components of a NetApp storage system to understand the performance and the capacity.
In this first chart, we demonstrate a balance chart showing the average, 10th/90th percentile and min/max values for processor busy. Similar charts show the average utilization for the disk drives in the aggregates. These allow you to know if you are close to the edge on your NetApp systems and whether or not you need to consider upgrading your configurations to handle additional load.
The second chart shows an overview of the Capacity usage. From here, we provide detailed capacity drill downs into the different components of the capacity.
SAN Switches: Brocade and Cisco
IntelliMagic Vision collects detailed error information from your Brocade and Cisco SAN switches and using this information, provides a rated view over time to provide an enterprise view of the health of your switches with the ability to drill down to the individual ports having the issues.
VMware: Drill Down from Application to Back-end
IntelliMagic Vision supports drill downs from VMware Guests to their associated datastores. You can also see the related volumes on the Datastore to see what other VMs are accessing. We also support performance views of the VMs so that you can understand who on the VM’s are doing what activity and isolate problem children.
Understanding Platform Specific Nuances in Multi-Vendor SAN Infrastructure
The examples I covered in this blog are only a small sample of the nuances each platform has on a distributed system. Understanding these nuances is crucial to providing intelligent and detailed performance management solutions that help you not only resolve issues, but prevent them.