Trend reports are designed to show you various storage trends over time.
...
At the bottom of many trend graphs is a selector that allows you to zoom in on a particular time period so you can focus on the data. The same kind of focus can be achieved by selecting specific dates from the date selectors or clicking one of the pre-defined zoom buttons. See our section on Navigating Your Data for more information on how you can interact with charts and graphs.
Enterprise Capacity Trend
This report is designed to show you six metrics graphed over time that relate to your overall enterprise storage.
...
It is fairly common for the used and allocated lines to be nearly identical. If you only see three of the six lines, remembering snap and orphaned values should be near zero, try turning off the allocated line to see if the used line is underneath allocated.
Capacity by Data Center Trend
This is a variant of the Enterprise Capacity Trend report. All the explanations for the graphs offered in the documentation of that report hold true for this one as well. The difference is this one is broken down by device type. The customizable data grid shows a selector for picking the device type you’d like to see. By default, it is set to the first device type in the list.
...
Changing the data center will alter the data displayed in the customizable data grid, and in turn, the data displayed in the line graphs above the grid.
Capacity by Device Type Trend
This is a variant of the Enterprise Capacity Trend report. All the explanations for the graphs offered in the documentation of that report hold true for this one as well. The difference is this one is broken down by device type. The customizable data grid shows a selector for picking the device type you’d like to see. By default, it is set to the first device type in the list.
...
Changing the device type will alter the data displayed in the customizable data grid, and in turn, the data displayed in the line graphs above the grid.
Capacity by Classification Trend
This is a variant of the Enterprise Capacity Trend report. All the explanations for the graphs offered in the documentation of that report hold true for this one as well. The difference is this one is broken down by classification. Classifications are set up when your organization is on-boarded with VSI. They allow you classify the different ways your organization uses storage. For example, you might have a classification for Backup storage, another for file share storage, and another for tracking your major applications like Oracle or SAP. This report allows you to review how storage is used across those classifications. The customizable data grid shows a selector for picking the device type you’d like to see. By default, it is set to the first device type in the list.
...
Changing the classification will alter the data displayed in the customizable data grid, and in turn, the data displayed in the line graphs above the grid.
Capacity All Data Centers Trend
This is a variant of the Enterprise Capacity Trend report. All the explanations for the graphs offered in the documentation of that report hold true for this one as well. The difference is this one is broken down by device type. The customizable data grid shows a selector for picking the device type you’d like to see. By default, it is set to the first device type in the list.
...
You can change the device type displayed in the graph by changing the selection in the drop-down list labelled Select Data Center.
Capacity All Device Types Trend
This report shows you a trend line for each data center within your global IT environment. You can change the metric you are viewing to any of the following:
...
| Note |
|---|
If you have a high capacity device type, it can sometimes dwarf the rest of the results making them un-readable. When that happens, turn off the offending device in the graph controls. For example, below we see an enterprise six device types, but the COHESITY_BACKUP device is significantly larger in terms of capacity and utilization than all the rest, so you can’t see the trends on the other devices. Turning off COHESITY_BACKUP using the line graph controls let’s us see the real trend on the other device types. |
...
Capacity All Classifications Trend
This report works the same way Capacity All Data Centers works. You have the same ten metrics:
...
As with the other reports, you can change the metric viewed for the report using the drop-down selector labelled Select a Space Type. This particular graph, like Capacity All Device Types Trend, shows a large skew in the data. The devices used for backup are significantly bigger in terms of capacity and utilization so their lines dwarf the other systems. You can turn those lines off by clicking the elements in the legend (see above for a visual example of this).
...
Performance Trend Reports
Most of the second half of the Enterprise Storage Trends menu focuses on performance. Many of these reports have a lot in common with each other, so let’s first look at some commonalities.
Each trend report is designed to show you two sets of data taken from two important sources. You’ll notice each chart has two Y axes. The positioning and color for these charts is intentionally the same on all our reports. The blue lines represent Input Output Operations Per Second (IOPS) which is measured as a count. We’re counting the number of operations that happen every second. The solid line represents your overall statistics for performance across your enterprise. The dashed line represents the same statistic, but focuses on your tier 1, your fastest storage. We would always expect the dashed line to be well below the solid line for the same metric. The corresponding Y axis is on the 👈 left side and is labelled IOPS. Again, IOPS is a count, so you are looking for numbers here in the hundreds of thousands.
The second set of lines in green represent the other measure for performance, your system’s latency. Latency is the average time it takes for an I/O request to complete. These statistics are graphed with green lines. As with IOPS, the solid line represents your overall latency and the dashed line surfaces your tier 1 performance. The Y axis for this set of statistics is on the right side 👉 , and its labelled using a measure of time in milliseconds.
...
| Info |
|---|
Users living with color blindness may have trouble with the colors on some of these graphs since given they depict the statistics in green and blue. If this is you, remember you can turn the lines on and off individually by clicking the lines in the graph’s legend. We have plans in the next major release to provide more support for these users in the next major release of VSI. |
Enterprise Performance Trend
This report focuses on your overall performance throughout your storage enterprise. See the above description regarding the meaning of the fields in the graph.
...
Field | What It Means | How It’s Computed |
|---|---|---|
Collection Date | The date the data were collected | Captured automatically when we run data collection |
IOPS | Input / Output Operations Per Second (IOPS) is a performance measure that focuses on the counted number of operations that can be completed within one second. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Tier 1 IOPS | This refers to the IOPS (see above) statistics with a focus on tier 1 storage, which is your fastest storage used for your mission critical and most demanding workloads. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Latency | Latency measure the time it takes to complete a single I/O operation. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Tier 1 Latency | This refers to the latency statistics with a focus on your tier 1 storage, which is your fastest storage used for your mission critical and most demanding workloads. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Performance by Data Center Trend
This report focuses on your overall performance throughout your storage enterprise. See the above description regarding the general meaning of the fields in the graph. This one doesn’t report your tier 1 performance separately, so you won’t see dashed lines on this graph. The IOPS reported here are peak IOPS rather than average which is to say we average the performance peaks in your data rather than presenting a weighted average of each device's own average.
...
Field | What It Means | How It’s Computed |
|---|---|---|
Collection Date | The date the data were collected | Captured automatically when we run data collection |
Data Center | The data center under review | You select this (see above) |
Latency | Latency measure the time it takes to complete a single I/O operation. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Peak IOPs | Input / Output Operations Per Second (IOPS) is a performance measure that focuses on the counted number of operations that can be completed within one second. In this case we’re focused on Peak IOPS which represents the usual IOPS statistic while the array is under load. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Performance by Device Type Trend
This report allows you to focus analysis of performance to a single device type. Use it to find out how well your NetApp, IBM, EMC, or whatever individual device type your interested in is performing. Further down the menu, we have a report that shows all these together in one report if you’re interested in a side-by-side comparison. You can change the device type using the selector labelled Select Device located just above the customizable data grid.
...
Field | What It Means | How It’s Computed |
|---|---|---|
Collection Date | The date the data were collected | Captured automatically when we run data collection |
Device Type Description | Shows the device type under consideration | You select this (see above) |
Latency | Latency measure the time it takes to complete a single I/O operation. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Peak IOPs | Input / Output Operations Per Second (IOPS) is a performance measure that focuses on the counted number of operations that can be completed within one second. In this case we’re focused on Peak IOPS which represents the usual IOPS statistic while the array is under load. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Performance by Classification Trend
This report allows you to focus your performance analysis on a particular classification of devices. This allows you to review the performance by some classification like “block storage” or “file storage”. The classifications are defined in your organization’s business unit data.
...
Field | What It Means | How It’s Computed |
|---|---|---|
Collection Date | The date the data were collected | Captured automatically when we run data collection |
Classification | Shows the classification under consideration | You select this (see above) |
Latency | Latency measure the time it takes to complete a single I/O operation. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Peak IOPs | Input / Output Operations Per Second (IOPS) is a performance measure that focuses on the counted number of operations that can be completed within one second. In this case we’re focused on Peak IOPS which represents the usual IOPS statistic while the array is under load. | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
...
The ALL Performance Reports
The reports in this section break away from the conventions used so far. Lovingly referred to as “The All Performance Reports”, these next few on the menu are designed to focus on the big picture versus the last set that focused on individual grouping metrics. So it’s nice to see performance by data center, but it might be equally or more useful to see all your data centers, or all your device types graphed together. In these reports instead of selecting a grouping category like data center, device type, classification, etc. you instead pick the performance metric allowing you to switch between IOPS, Latency measured in milliseconds, and a new (or at least different) metric Megabytes Per Second (MB/s) which is a measure of overall throughput. Like latency it is a measure of performance related to time, but MB/Sec is often a more practical metric since it focuses on how much actual data is moving to and from the device rather than focusing on the average timing of a single request.
...
Remember for these reports we’re not showing both axes, we’re only showing one which you can pick using the selector. IOPS shows up first alphabetically, but you need to check your latency as well as your IOPS numbers. Furthermore, the colors are essentially random on this graph, since they represent organizationally specific data, so focus on the legend at the bottom.
Performance All Data Centers Trend
This report shows all your data center status in a side-by-side comparison. This will help you identify any data centers that are underperforming relative to your organization’s performance standards. Your data centers are listed in the legend beneath the the graph. There is one line per data center and they are presented in an essentially random color scheme. In the reports earlier that focused on these measures one at a time, for example one data center at a time, the colors had specific meanings, here they don’t.
...
Remember, the colors here are random. If you see a bunch of red lines it doesn’t necessarily indicate a problem.
Performance All Device Types Trend
This report allows you to do a side-by-side comparison of your performance based on your device types. This helps you find your fastest arrays by type, and it can help you identify underperforming storage based on those types. For example, we would expect a Pure Storage array consisting of entirely flash storage to consistently outperform a legacy NetApp FAS device which is running 10K SAS drives. If you were to see a flash system with a line closer to a traditionally slower system’s line, you know something is wrong.
...
Remember, the colors here are random. If you see a bunch of red lines it doesn’t necessarily indicate a problem.
Performance All Classifications Trend
This report shows you a side-by-side comparison by classifications. This will allow you to review how well your various classes of storage are performing relative to one another. For example you’ll be able to see how your block storage performs in comparison to your file storage. These classes typically have different performance characteristics, so you’re looking for unlikely convergences.
...
You know that drawer everybody has in their kitchen that houses all the important stuff which doesn’t seem to belong anywhere else? Like batteries and tape. You need them, but you don’t need a battery drawer the way you need one for your flatware. Welcome to that drawer in this menu. These next two reports show vital information but don’t fit the molds presented by their predecessors.
Tier by Data Center
This report shows you a break-down of capacity based on storage tiers within each data center. You select which data center you want to view using the selector above the customizable data grid. The lines on this single axis graph represent the different tiers defined within your business unit data. Most enterprises use a three tier system with your fastest (and usually most expensive) storage belonging to tier 1. Tier 2 is your intermediate in terms of performance and cost per gigabyte, while tier 3 typically represents the slowest and cheapest storage in your inventory.
...
Field | What It Means | How It’s Computed |
|---|---|---|
Collection Date | The date the data were collected | Captured automatically when we run data collection |
Data Center | Shows the data center for that line of data | This is defined within your organization’s business unit data |
Tier 1 Free | The amount of free space on each collection date for tier 1 storage | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Tier 2 Free | The amount of free space on each collection date for tier 2 storage | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Tier 3 Free | The amount of free space on each collection date for tier 3 storage | Your devices report this statistic directly during data collection. VSI uses a weighted average for summarizing your data on this graph. |
Multi-Tier Free | The amount of free space on each collection date for storage that is defined within multiple tiers. | |
Tier 1 Used | The amount of used space on each collection date for tier 1 storage | |
Tier 2 Used | The amount of used space on each collection date for tier 2 storage | |
Tier 3 Used | The amount of used space on each collection date for tier 3 storage | |
Multi-Tier Used | The amount of used space on each collection date for storage storage that spans the tier definitions. |
Red Light Alert Trend
With each data collection you’re getting a new set of alerts. This report shows you the trend for those alerts over time. You can often link spikes in this report to specific events like maintenance, upgrades, adding storage to an array, etc. The trends here allow you to correlate these kinds of events with spikes in the alerts so you can improve your processes in the future to account for the spikes.
...