Predicting the capacity of a network without reliable data is a risk to many companies. In fact, many companies still anticipate capacity needs based on historical usage trends, the use of spreadsheets, and the use of separate monitoring software. While all these methods provide some degree of insight into usage, they do not give a comprehensive view of what is happening on the network, how it is used, and what capacity it may require moving forward.
This article explains how network inventory facilitates accurate capacity forecasting, the limitations of historical forecasting methods, and how modern enterprises use network inventory data to create better capacity forecasts, mitigate risk, and align investments in IT infrastructure with the growth of the business as per Slurp’it.
Why has capacity forecasting become more complex?
In the past decade, enterprise networks have been relatively stable. Networks were confined to local premises; application life cycles tended to be lengthy, and user access followed established patterns.
Thus, capacity planning was typically only performed once or twice per year when additional bandwidth or hardware needed to be added. Traffic flows have become less predictable, leading to significant fluctuations in daily traffic patterns as per Slurp’it.
As per Cisco’s Annual Internet Report, the increase in global IP traffic can be attributed mainly to the increasing number of cloud applications and increased video-related workloads; however, by 2026, Gartner estimates that over 75% of workloads in enterprises will be running in cloud or hybrid environments.
Now what are the problems of traditional capacity planning?
Unfortunately, many organizations rely only on monitoring tools as the basis for their capacity forecasting. While monitoring tools do provide information about performance, they lack context. The problems are listed below:
- An alarming example would be that of a monitoring tool displaying that there is an increase in the bandwidth utilization of a link.
- If, however, you do not have the inventory data to see what devices are physically connected, you may not know which applications are driving the increased traffic or if this increased traffic is a normal growth pattern or just an anomalous event.
- Using spreadsheets as a capacity planning Tool also acts as a barrier since they require the owner to routinely update his or her spreadsheet.
- Because of the dynamic behavior of networks over time, manual updating of performance data and resource consumption to an updated spreadsheet will often result in inaccurate or untrustworthy data.
How historical inventory data could help in identifying growth trends
Capacity forecasting considers an organization’s historical inventory data and how the organization’s network has developed.
- To effectively manage and plan a network’s future, historical context must be kept in a properly maintained inventory of a network’s inventory.
- To best understand how a network will grow, a network’s historical record of when each asset was added, modified, or removed will provide insights into the growth trends of a company’s network.
- For example, the historical inventory of data may show that the number of cloud workloads has doubled in the past 12 months and that the east-west traffic in the cloud environment is currently growing at a faster rate than the north-south traffic.
These insights allow organizations to project and prioritize capacity upgrades based on their forecast of what, when, and where their future capacity needs will be.
What will happen if forecast accuracy increases around remote and distributed workforces?
Work-from-home trends have created new patterns of network usage. VPNs, secure access gateways, and cloud applications are all contributing to demand fluctuations.
Using network inventory enables organizations to understand how their remote access infrastructure is deployed and used, including the number of gateways customers are using and how many users connect through each, and how traffic is routed into the core systems of an organization.
The visibility provided by the use of it will assist organizations to more accurately forecast remote access capacity, especially during peak usage times.
Additionally, network inventory will identify potential single points of failure that could impact a high number of users due to capacity shortages.
As remote and geographically distributed workers become the norm, network inventorydriven capacity forecasting will be a fundamental element for maintaining end-user performance and experience while working remotely.
How risk can be reduced through proactive capacity forecasting
By having a proper network inventory, network teams will be able to make proactive decisions regarding inventory forecasting to prevent capacity shortages.
For example, through network inventory, organizations will be able to identify which devices are approaching their maximum capabilities, e.g., the number of years a device has been in service is directly related to performance degradation. It will also assist in identifying the number of links that consistently hit the maximum capacity.
Thus, with the information provided by network inventory, the network IT department may effectively establish upgrade priorities based on risk and business impact instead of waiting until the day after a critical failure as per Slurp’it.
Some best practices
Organizations should follow a few best practices for using it effectively. Let us see which are those in the following:
- Accurate and up-to-date inventory data is a must when conducting capacity forecasts based on current network usage patterns by leveraging a combination of both network-stimulated and cloud-stimulated data-points.
- An organization must be able to constantly track how and where network data items are used in order to optimize resource usage and revenue streams associated with those resources.
- A second best practice is to monitor inventory against usage data that is linked to specific assets, as inventory visibility enhances the quality of capacity forecast.
- Using both cloud and virtual datacenter inventory alongside one another provides more complete inventory visibility, resulting in more accurate capacity forecasts as per Slurp’it.
- An additional best practice for optimizing the value from inventory visibility is to maintain a history of inventory usage and volume for all inventory items.
Lastly, in addition to maintaining an accurate inventory, organizations should engage both their application and network teams in capacity forecasting via a cross-functional forum. Inventory visibility creates a common language between these two groups, providing an opportunity for enhanced collaboration when conducting capacity planning discussions. For more information, contact us at Slurp’it.