New Linden Lab VP with an old message
Linden Lab's Frank Ambrose has weighed in with his first public statement. Unless you've been following the Second Life news here at Massively, you probably had no idea who Ambrose is, or that he'd been hired by the Lab as Senior VP of Global Technology.
Ambrose delivers the basic message that we hear in virtually every infrastructure-related announcement for the last several years: Scalability and stability. Unfortunately, after three years, users indicate that they're finding the promises wearing a little thin.
That is not to say that there haven't been some improvements in these areas. On Sunday, 14 September Second Life user-concurrency rose to 69,754 (and notably without an associated infrastructure collapse). That's an overall increase of 17,859 in one year. If that trend continues, Second Life will have eclipsed the popularity of Everquest at its height, in a little more than a year.
 | Are you a part of the most widely-known collaborative virtual environment or keeping a close eye on it? Massively's Second Life coverage keeps you in the loop. |
Ambrose highlights two areas for near-term improvement. The first being the already extant project to migrate infrequently-accessed assets out of the high-performance database clusters. Currently the asset clusters are weighed down by vast numbers of asset resources that are rarely ever required by a simulator or viewer.
Presumably this will entail slow access to these uncommon resources, but it isn't really clear how users would even notice a speed reduction in asset access, unless it causes a significant improvement in the speed with which 'fast' assets are fetched and delivered.
The second item, though, is a new one.
Ambrose is targeting the VPN links between systems and facilities. If you're a frequent Second Life user with an eye for outage notices, the merest mention of the acronym VPN (for Virtual Private Network) probably causes you to cringe in anticipation of widespread infrastructure and services failures.
Certainly the VPN lashups have been responsible for a considerable number of failures. Primarily they're used in order to keep data moving between systems private, by encrypting server communications that have to be hauled across third-party networks -- you don't want Level 3 reading all your IMs, right?
However as a frequently failing system, it is obvious that something has to be done -- and that will probably involve some new form of encrypted inter-server data transport. Putting our heads together, we can see no less than four alternative solutions that are cheap and rapid to implement. Actual deployment when the time comes, may require taking the whole grid offline, so we're hoping that considerable testing will be employed to ensure that whatever solution is chosen really does work better than the existing VPN lashups.
At the end of the day, though, what is highly visible among users is message-fatigue. Linden Lab employees mentioning scalability and stability seems to be having an increasingly soporific effect among an increasingly dismissive user-base.
Popular digital doyenne Ordinal Malaprop, perhaps summed that feeling up most succinctly, 'In any case: please wake me when things work. I have a little sign around my neck, it is quite obvious.'
