With the announcement of Lync 2013 (previously known in the beta circles as Lync 15), many are interested in what's new. However in a nutshell, Lync 2013 doesn't re-invent the wheel when compared to Lync 2010. Lync 2013 builds on the features introduced in Lync 2010 in a way that makes Lync2013a compelling upgrade. The following is a rundown of the new main features:
There has been significant role consolidation in Lync 2013 because now you no longer have a separate server role for monitoring and archiving. Each Front End server communicates directly with the monitoring and/or archiving database, eliminating the need for a separate monitoring/archiving server.
As far as the A/V conferencing server role, you can no longer install the A/V conferencing server role separately. It is now “always” co-located with the Front End role.
Directors are now an optional role. The Director role is no longer "recommended" but optional; Microsoft says you can "safely exclude the Director with confidence that the Front End Servers will provide the same services in their place."
There is a new server role for Persistent Chat, formerly known as group chat. Persistent Chat will be discussed in more detail to follow.
DR/High Availability Options
Lync 2010 introduced the concept of a backup registrar. When a user's home pool becomes unavailable, the client can automatically register with a pre-defined backup pool. This maintains basic voice availability, but the client loses conferencing capabilities, and the user's contact list is unavailable. In Lync 2013, users maintain nearly all functionality in the event of a failed pool. This is made possible because all user data is now replicated between all Lync servers in the enterprise through pairing of pools and is supported across datacenters with “no” restriction on the distance between datacenters. Additionally every server maintains multiple copies of the user database, so there is almost no reduction in service availability, "almost" because Response Groups are still not highly available (something that was sorely missed in Lync 2010). So should a failure occur on a home pool that hosts Response Groups, those response groups will not be available.
Each Front End server stores a complete copy of all the databases stored in the SQL back-end, so if the back-end SQL database server is unavailable, the Front End will still function. Additionally Lync 2013 supports SQL mirroring on the back-end databases. This can reduce hardware costs typically associated with the older clustering options in SQL such as separate shared storage.
A little more on high availability and resilience will be discussed under Load Balancing to follow.
In Lync 2010, if you had multiple mediation servers connecting to the same PSTN (Public Switched Telephone Network) gateway or SIP trunk, you had to fake the Topology Builder out by creating multiple DNS “A” records pointing to the same IP. Lync 2013 now supports M-N trunk routing. This allows you tohave multiple trunks to different gateways, and a gateway to have multiple trunks to different Mediation Servers.
Lync 2013 includes support for inter-trunk routing. This feature allows Lync to act as an intermediary between two or more different phone systems. For example, Lync can accept calls from one PBX, and pass the call through to another PBX. This can be very useful in larger environments and allows Lync to be the backbone of a corporate telephone network.
In Lync 2010, you could use trunk translation rules to modify the CALLED phone number before passing it to the next hop. However, you couldn't make any changes to the CALLING number (ie: the person making the telephone call). Lync 2013 now allows you to make changes to both the called and calling number. This is very useful when the PSTN provider does not accept E.164 formatted phone numbers. For example, in North America, many PSTN providers do not accept the country code 1 as part of the number and only accepts 10-digit numbers. In the past, an external gateway would have to do the necessary manipulation, but with Lync 2013, all the number manipulation can be done within Lync.
There are also several other new Enterprise Voice related enhancements. Delegates can setup simultaneous ringing to their mobile devices for incoming calls to their manager. When a user has setup simultaneous ringing to a mobile phone, and the device is turned off or out of range, Lync 2013 can determine that an incoming call was immediately routed to voicemail, and disconnect that endpoint so the call can continue to ring other endpoints. Caller ID presentation allows administrators to modify the Caller ID format in a much more scalable way than in Lync 2010, which only allowed Caller ID changes based on the route.
Integration with Lync Online
You can now create hybrid deployments with a mix of on-premises and Lync Online servers (similar to Exchange 2010). This means that you can have some users running "in the cloud" and some users on traditional on-premises servers. Microsoft calls this "hybrid voice". You can also have all your users running in Lync Online and make calls via an on-premises PSTN gateway. This means you can allow Lync Online users to dial legacy PBX extensions, or make calls via a traditional PSTN connection (T1/E1 or similar) in situations where SIP trunking isn't desirable or an option. Media bypass will work in this situation, so a user's media stream won't be hair pinned through the Lync Online service when making phone calls from an office running a local PSTN gateway so no doubt more integration flexibility built into Lync 2013.
Persistent Chat (or group chat) as was mentioned above, is now a full-fledged Lync service or role, unlike older versions which was really just tacked on. You can now define servers in the Topology Builder as with other roles, and the Persistent Chat features are included in the base Lync 2013 client as no separate client is required. This new server role has several components: PersistentChatService, PersistentChatStore, and PersistentChatComplianceStore. The term "Persistent" in this new role means that a history of the chat session is retained, so users can jump into a chat room and get up to speed with the conversation that has already taken place. Some people suspect this could be a replacement for email distribution lists by providing a constant, updated message repository, like a Wiki for example.
Other New Neat Features in Lync 2013
Other features that don't fall into the above categories include:
- Full A/V capabilities on the Lync Web App client.
- Mobility - the ability to make audio and video calls from a mobile device using either a mobile data or Wi-Fi connection.There are many Lync 2013 mobile clients available, visit this link for more detailed information: Lync 2013 Mobile Clients Release: FAQ.
- IPv6 is fully supported, in addition to support for IPv4. So you can go dual-stack (IPv4 and IPv6) if desired.
- VDI plugin - allows full A/V support in virtual desktop environments.
- Support for HD resolution up to 1080p. Also, the H.264 video codec is the default for video encoding, which brings with it greater flexibility for resolutions, frame rates, and scalability.
- Skype Federation Support.
- A Unified contact store with Exchange Server 2013, as well as the ability to use high-resolution photos for contacts.
- Archiving integration which enables Lync to work with the Legal Hold feature in Exchange for compliance and e-discovery needs.
- Minor changes to Response Groups, basically you can configure Response Group Managers and Administrators in Lync 2013, allowing you to delegate Response Group tasks to other users. This feature was in OCS 2007 R2, but was removed from Lync 2010 for some reason.
- Client compatibility.Lync 2013 expands its client compatibility to include IE10, Windows 8, Windows Phone, iOS, and Android.
There are a lot of other small enhancements that go a long way towards improving the overall product or enhancing usability but without a doubt these are the most significant improvements from Lync 2010 to Lync 2013.
As far as Load Balancing is concerned, even though there have been significant enhancements to the Lync Server architecture for improving availability, not much has changed from Lync 2010 to Lync 2013 from a Load Balancing standpoint however there are a few items worth mentioning and should be taken note of.
As mentioned above, Lync 2013 includes new feature enhancements such as paired pools and SQL database mirroring which provide a robust and effective framework for maximizing uptime and these availability enhancements will help to meet even the most stringent service-level agreements (SLAs). However along with these enhancements, the GEO LoadMaster appliance made by Kemp Technologies can assist with automating the distribution of traffic from one site to another while ensuring seamless failover and failback to the best performing and geographically closest datacenter thus optimizing “up time” of Lync web traffic.
One example of how the GEO LoadMaster has proven to be very useful with Lync 2010 is by providing site resilience for the simple URL’s (meet, dial-in, admin) however with Lync 2013, because of the high availability enhancements built-in an example of how the GEO can prove to be very useful is with two Edge Server/Pool deployments (one per site) and both are using a single Access Edge FQDN such as sip.contoso.com. These are not the only scenarios where the GEO LoadMaster can be used and provide more automation when it comes to site resilience, basically if it is web traffic that needs to be made resilient between multiple datacenters, the GEO is the perfect solution. In a nutshell, what the GEO does is provide a response to a DNS query for a particular FQDN and sends the requester an “A” Record which resolves to the server host/pool IP or optimally, the internal LoadMaster VIP address used for Load Balancing the traffic to the server pool. More detailed information regarding the GEO LoadMaster can be found through the following links:
In regard to Load Balancing Lync traffic within a particular datacenter or site, when Load Balancing Lync Client Mobility with Lync 2010, the only option was to use a Layer 7 Virtual Service and Microsoft recommended using Cookie Persistence as the preferred persistence method. The cookie to be used was the “MS-WSMAN” cookie issued by the Load Balancer so on the Kemp LoadMaster, the persistence mode should be set to “Active Cookie” with a fairly high persistence timeout. See the screen shot below as an example of how the configuration would look like on the LoadMaster.
Figure 1 – Example of “Active Cookie” configuration on a Kemp LoadMaster.
With Lync 2013, Microsoft’s recommendation is to use either a Layer 4 or a Layer 7 Virtual Service. If you use Layer 4, the recommended persistence method that should be used is “Source IP” Address. The reason for the flexibility is because the requirement for cookie affinity or persistence has been dramatically reduced, and you can provide a substitution by using Transmission Control Protocol (TCP) affinity if you are using Lync Mobile delivered with Lync Server 2013. Cookie affinity can still be used, but the web services no longer require it. Other than that there are no other differences with Load Balancing between Lync 2010 and Lync 2013. Please visit the following links for more details and examples on Load Balancing Lync Mobility.
To conclude, although Lync 2013 does not re-invent the wheel when compared to Lync 2010, there are certainly many new desirous features and advantages to upgrading especially when it comes to the topology changes, high availability and disaster recovery. Paired Pools provide site resiliency without placing limitations on distance between pools. More importantly, they ensure that all workloads are truly resilient and with these features in place along with the proper Load Balancing solution, Lync Server 2013 offers a near-seamless end user experience in the event of a failover.
Another compelling reason to consider upgrading is the fact that not much has changed when it comes to Load Balancing requirements so following the same guidelines that were set out for Lync 2010 with regard to configuration, the number of Load Balancers needed whether a hardware or virtual appliance is used, and where to deploy the Load Balancers in the network infrastructure will work just as great for Lync 2013 as it did for deploying Lync 2010. Please visit the following links for more detailed information: Lync 2010 Deployment Guide.
Furthermore the Web User Interface (WUI) on any Kemp LoadMaster appliance (hardware or virtual) makes deploying either Lync 2010 or Lync 2013 really simple. This can be seen firsthand by either downloading a virtual trial from this link: Kemp's Virtual Product Matrix or contacting Kemp Sales to sign up for a hardware try and buy.
Last, once you have our product in hand whether a hardware or virtual appliance, the Kemp Support Team is made available for assistance on a 5 x 10 basis. This includes assistance with installation and/or deployment of Lync 2010 or 2013. More detailed information can be discussed with a Kemp Sales Representative. Please visit this link to contact Kemp Sales: Contact Kemp.
That said, anyone who deploys Lync (2010/2013) using Kemp Load Balancers will truly echo the words of the feline below…