Microsoft’s SharePoint 2010 has become one of the more important tools for sharing information across the enterprise. SharePoint’s live collaboration capabilities and document management tools have increased productivity for workgroups and ad-hoc teams across many businesses.
The rapid adoption of SharePoint 2010 and the growth of SharePoint’s managed data sets is a true testament to its value. However, many organisations are discovering that as its SharePoint deployments grow and spread across multiple branch offices, performance suffers.
While that is true of most any technology deployed across multiple sites, the effect seems more noticeable with SharePoint, simply because of the products ability to work in “real-time” environments. Simply put, any time an end user is affected by a performance issue, productivity takes a hit.
The traditional methodologies for improving performance usually consist of throwing more processing power at the problem – with SharePoint that would mean more servers, faster storage and a significant investment in hardware. However, that only scratches the surface of the real problem, which turns out to be a connectivity related issue, an issue that is defined by bandwidth, throughput and latency, all of which are beyond the control of the SharePoint server.
Nowhere are those performance issues more evident than with branch office connections and remote or mobile workers.
Addressing performance problems for large, distributed SharePoint teams requires a different approach, an approach that is not driven by server processing power and increased bandwidth, but an approach that adds intelligence into the infrastructure.
This issue here is that workgroups (or team members) are encountering reliability and performance issues as the size of the team grows and as the amount of data managed increases. That proves to be a real conundrum. SharePoint’s value increases as more data is managed and shared, but that value lessens as performance degrades, which is a result of adding more data, more sites, more users and more applications.
Most enterprises try to solve the problem by scaling up SharePoint servers, adding SharePoint servers to branch offices and incorporating complex asynchronous data replication routines.
The net result is an increase in wide area network (WAN) traffic, which ultimately negatively impacts overall performance. However, there are other solutions to that performance problem - solutions that are ultimately less expensive to deploy, improve performance, reduce latency and harden security, all without attempting to decentralise a SharePoint presence by installing more SharePoint Servers at more locations.
It all comes down to WAN optimisation. By incorporating these solutions, performance can be enhanced greatly, sometimes by a factor of 100 or more, thanks to a reduction of redundant requests, effective data compression, reduced latency and packet de-duplication. However, all WAN optimisation schemes are not created equal, there are certain capabilities that provide more benefits than others.
Applying WAN optimisation technologies to a SharePoint environment takes more than just dropping a WAN optimisation appliance into the mix. Administrators will need to identify where the bottlenecks are and what technologies will best address those bottlenecks, while enhancing the end user experience. Ideally, the goal is to provide a LAN-like experience to remote office and mobile SharePoint users, and that may take a little research to figure out what technologies will offer the biggest bang for the buck.
Perhaps the first subsets of users to address are the mobile users. Users that may access SharePoint from a customer location, home office, Wi-Fi hotspot, hotel VPN or any other remote connectivity technology. Those users prove to be a unique challenge – because of the mobility involved, it is difficult to automatically assign those users to a particular branch office or WAN entry point.
Here, a solution that leverages mobile WAN optimisation may prove to be best. With mobile WAN optimisation, a small client application is installed on the user’s notebook or remote PC. That client software connects via the Web to a WAN optimization appliance located in the data center or a branch office.
The appliance and the client perform some handshaking and determine the fastest way to transmit data over the WAN. Normally that includes a combination of data compression, data de-duplication, algorithms to reduce chatty protocols and a measure of data caching, which eliminates unnecessary data retransmissions.
A mobile WAN optimisation solution can offer major performance improvements. For example, local traffic caching can reduce the amount of synchronisation traffic significantly, while improved compression reduces the size of the traffic being physically sent, from those two technologies, users can expect anywhere from a five to fifty fold increase in transmission performance, simply because SharePoint uses very “chatty” synchronisation schemes which incorporate hundreds of round trips for packets.
When those round trips are reduced and chatty protocols are tamed, the performance improvement is dramatic.
In some cases, it may be best to look at a distributed architecture for SharePoint, especially for businesses that have a multitude of large branch offices. In that case, WAN optimisation appliances can be deployed at the edge of each branch office and the data centres. Those WAN optimisation appliances can auto-discover each other and then can monitor and determine the best paths for WAN traffic.
Occasionally, that may involve asynchronous relationships; otherwise, it may incorporate synchronous relationships between the sites. Ideally, multiple WAN optimisation appliances will be able to automatically determine those relationships and adjust them on the fly. However, there are additional benefits offered in a multi-site, multi-appliance scenario, the first being resiliency, as wells the ability to support multiple mobile users from each branch office by pairing a mobile WAN optimisation client with each appliance.
For those looking to meet business continuity needs, WAN optimisation appliances can enable SharePoint servers to become more resilient, where if one server or site fails, traffic can be rerouted to the next nearest site. High speed synchronisation powered by WAN optimisation keeps the data up-to-date.
What’s more, multiple appliances have the ability to incorporate highly compressed links between sites and incorporate intelligent caching of information, keeping the most frequently used data packets close to local users. Advanced algorithms can also pre-fetch data, allowing local caches to provide the information before a workstation requests it from the remote site. Other advantages include increased security – appliances can encapsulate, encrypt and compress the WAN traffic between sites reducing latency and increasing throughput.
SharePoint responds very well to optimisation technologies, perhaps to the point where SharePoint servers may not have to be installed in the branch offices and the WAN optimisation appliances handle all SharePoint traffic from the data center to the workstation and still offer LAN-like performance.
The moral of the story here is that it may not always be necessary to turn to complex scale outs with servers scattered about different geographic locations to provide acceptable performance to end users, especially with SharePoint. Alternatives, such as WAN optimisation can often offer the needed speed, without requiring local SharePoint resources, and ultimately, WAN optimisation appliances will cost less than branch office servers, while offering to improve all WAN traffic and not just be limited to optimising a single application or service.
WAN optimisation works to improve the end user experience, regardless of the user’s location. That user can expect a LAN-like experience while on the road, seated in a branch office or from a home office, thanks to WAN optimisation technologies.
Mark Lewis is Senior Director Marketing and Alliances EMEA at optimisation company Riverbed Technology