|
Threads
View context
That's a fair question and I can only speculate an answer as I'm not an admin but a regular user that offered myself to be the link between admins with very little time in case of disruptions like this one.
While very similar on the surface Open Sim and Second Life are very different. Someone said OS is an open source fork of SL and I believe it was at first but their technology evolved divergently mostly because money. (Again, this is pure speculation on my part).
We don't have the same levels of resources SL has, this is an open source, volunteer driven project of very few people putting an unholy amount of hours for free just for us to have a place to play around.
Also OS situation is exponentially more complicated than SL.
Think that OS must keep a compatibility, backwards and forwards between multiple forks of itself, in that regard SL has it easy, it's just one closed grid. OS is more like a constellation of grids that sends and receives a huge amount of data constantly.
Think about an hyper jump, suddenly your inventory must be available in another grid, my guess is that the communications between grids, caches, databases, accounts, inventory, etc is just enormous.
So enormous in fact that the asset server(s) have to handle millions of operations every day, in spite of that we're very few users compared with SL the amount of data processed is staggering. With less resources than a paid service. This outage is to literally make more space in the asset server to recover the assets that were transferred from the old datacentre to the new one and to keep operating with the regular ingestion of new assets. We're talking about server grade hardware, for a very niche and specific application that has to be tuned and setup by one or two guys that makes this happen for free.
So to answer your question, I'm sure the approach is different (as far as I understand SL is putting the data in the cloud now), what's going on is also different (not only the space optimization but also the recover of the data from the old asset server), it's running on a shoe string budget by very few people that's doing it for free.
Oh btw, as far as i understand tomorrow things should be up and running again, if something different happens i'll post it here.
Edit: And i did, I didn't realise the Tweet but I understand now the question. Technically is just Microsoft time if you move around 4 huge files it'll take way less time than if you move around 4 million tiny files that results in the same occupied space.
like(1)
While very similar on the surface Open Sim and Second Life are very different. Someone said OS is an open source fork of SL and I believe it was at first but their technology evolved divergently mostly because money. (Again, this is pure speculation on my part).
We don't have the same levels of resources SL has, this is an open source, volunteer driven project of very few people putting an unholy amount of hours for free just for us to have a place to play around.
Also OS situation is exponentially more complicated than SL.
Think that OS must keep a compatibility, backwards and forwards between multiple forks of itself, in that regard SL has it easy, it's just one closed grid. OS is more like a constellation of grids that sends and receives a huge amount of data constantly.
Think about an hyper jump, suddenly your inventory must be available in another grid, my guess is that the communications between grids, caches, databases, accounts, inventory, etc is just enormous.
So enormous in fact that the asset server(s) have to handle millions of operations every day, in spite of that we're very few users compared with SL the amount of data processed is staggering. With less resources than a paid service. This outage is to literally make more space in the asset server to recover the assets that were transferred from the old datacentre to the new one and to keep operating with the regular ingestion of new assets. We're talking about server grade hardware, for a very niche and specific application that has to be tuned and setup by one or two guys that makes this happen for free.
So to answer your question, I'm sure the approach is different (as far as I understand SL is putting the data in the cloud now), what's going on is also different (not only the space optimization but also the recover of the data from the old asset server), it's running on a shoe string budget by very few people that's doing it for free.
Oh btw, as far as i understand tomorrow things should be up and running again, if something different happens i'll post it here.
Edit: And i did, I didn't realise the Tweet but I understand now the question. Technically is just Microsoft time if you move around 4 huge files it'll take way less time than if you move around 4 million tiny files that results in the same occupied space.
View context
