Your storage lead time is the time is takes to allocate more storage to a host volume once you realize you require additional storage.  It includes the time it takes to put together your purchase order, get it approved, chose a supplier, ship the new disks, and then install and provision the new storage.  For many companies, this can take up to three months.  For slow growing volumes, this can be sufficient timing, but for high growing volumes, utilization can reach 100%, or become dangerously close to 100%, within three months time.  This is how fires get started.

An option to resolve this issue is to lower your utilization threshold. Lowering it would greatly reduce the probability that you will run out of storage on high growth volumes.  However, the drawback to this approach is that you will add storage to your numerous slower growing volumes way too soon.  This will reduce your overall storage utilization below the already low 30% - 40% levels seen in today's average company.

An efficient way to resolve the "when do I add more storage?" dilemma is to based your decisions on usage forecasts.  Advanced statistical algorithms can be applied to historical storage growth for each host volume and accurate storage usage forecasts can be developed.  Now we can schedule additional storage purchases when the forecast crosses the utilization threshold and then subtracting out the storage lead time.  This will ensure you avoid out-of-storage conditions on fast growing volumes without wasting storage by allocating it to your slower growing volumes too early.

Storage usage forecasts are the heart of any well-defined capacity planning process.  By evaluating how much storage will be consumed in the future, storage administration groups can:

• Get early warnings on future out-of-storage conditions far before they occur.

• Proficiently consolidate storage by identifying usage rates and areas of over-provisioning.

• Reduce storage CapEx spending by increasing the utilization of existing storage and future storage purchases.

• Intelligently configure new storage subsystems during migrations and technology refreshes.

• Understand the impacts of data protection schemes.

• Effectively plan for disk to disk backups.

When you see that the utilization for a number of host volumes will exceed your thresholds in an amount of time that equals your storage lead time, the next question you're faced with is "what is the efficient amount of storage that I should add?" to the host volumes.  You can increase efficiency by adding a small amount of storage, but you don't want to have to keep adding additional storage very frequently.  On the other hand, you can add a large amount and not have to worry about adding more for a long time, but this will also reduce your overall storage utilization below the already low 30% - 40% levels seen in today's average company.

A good rule of thumb is to look at your utilization forecasts and allocate enough additional storage such that the additional storage will last 2 times the duration of your storage lead time.  For example, if your storage lead time is 6 weeks, look at how much additional storage the application will consume in the next 3 months, and allocate that much more usable storage to that volume.  This will keep your storage utilization efficient while reducing the amount of times you need to add additional storage.

Today MonoSphere announced the newest release of its award-winning software, Storage Horizon™ 2.3.  The solution provides Global 3000 companies a unified view of current and projected storage capacity usage, which enables IT organizations to develop effective storage-capacity plans, maximize utilization of existing storage assets, and significantly reduce CapEx spending.

Storage usage can change significantly over time, making it necessary to frequently update usage forecasts to reflect new trends.  In large environments with tens to hundreds of thousands of volumes and filesystems, it is not feasible to manually generate all these forecasts.  To address this, MonoSphere's latest version of Storage Horizon features an Auto-Forecaster.  Through user-defined templates, end users can schedule periodic forecasts to run automatically across all their host volumes and supported applications.  With update-to-date forecasts, Storage Horizon's analysis allows users to keep a close watch on their entire storage infrastructure and get notifications of potential issues long before they become problems.

In addition, Storage Horizon's forecasting engine will automatically assign each forecast a Modeling Quality Index, which uses advanced statistical algorithms to calculate the quality of the forecast.  Storage Horizon categorizes each index as either: Poor, Good or Excellent.  A Modeling Quality Index rating of "Excellent" would be assigned, for example, if a six-month forecast was generated on 6 months of collected data.  When running reports from a Storage Horizon analysis engine, such as the Storage Planner, users can specify a Modeling Quality Index of "Good or Better" to avoid making poor purchasing decisions based on insufficient data.

Lastly, Storage Horizon 2.3 will add support for data collection and analysis from VMware® ESX.  As with previous versions, Storage Horizon 2.3 does not require installation of software agents on application servers, so the system deploys in hours and collects storage usage data immediately.