did you read the blog recently...

The problem is a combanation of two things 1) Massivly Huge Database 2) Poop


when they mix, well you know what happens....

I still wouldn't re-up. Instead of premiums, I think perhaps payment USED persons(Maybe a min amount too)?

Hard to know for sure. Force the Traffic and camperbots OFF the grid for a couple days and see what happens.

How do you do that? Only by painstakingly finding the Bot groups can you do that with any certainty. If you restrict logins to only Premium and Payment Info Used then a lot of people who fall into the following categories will be excluded too and may not come back: -

LOL its easier than you may think.

Just go to the highest traffic places on private ISlands and the Mainland and start /kick ing people

......

So I don't think that is a good idea - maybe pushing Premium people to the head of a login queue is a reasonable way of distinguishing Premium accounts but cutting off access deliberately for some of these people in my opinion would be a mistake.

UPDATE 4:09 p.m. PST - kate] In world services are mostly available. There are still intermittent reports of stale transactions.

The Dawning of a New Viewer: Second Life 1.19.1 Now Available!

Friends. Romans. Residents. After months in First Look and Release Candidate phases, we are pleased to announce that the Second Life 1.19.1 Viewer, featuring WindLight atmospheric lighting and gobs of other stability, performance, and feature enhancements, is now available for download! And so….
...
But what goodness does it contain?…

Premiums get priority, freebies join the login queue.

I'd prefer to see those with payment info on file getting priority.

I'd prefer to see those with payment info on file getting priority.

Philip's vision for SL is very egalitarian, very net-neutral. I'm not sure how fond he would be of ANY preferential access.
.

I would not be surprised to learn that a meeting went down where Phillip said "if you do this, I will resign as CEO"... and they said "fine".

If I were Philip and Robin, I think I'd stonewall on the question of preferential access to different classes of customers. I'm cynical, I guess, but deep in my pea brain I think I know how that discussion would go in Philip's staff meeting.

Robin: I have a proposal my people have worked up for tiered access based on account type. *presents something like what has been discussed on this thread*

CFO (chief financial officer): That's not the way to do it.

Philip: Oh?

CFO: Right. Preferential access goes to major accounts ... that's any Fortune 5000 corporate account, significantly sized nonprofits, strategic accounts (ESC, MOU), and individual accounts that have more than x sims. Everyone associated with those accounts, including their bots, get access. Then premiums. Then paid info on file. Then basics.

Philip & Robin: *Exchange unhappy looks*

CFO: Let me call Bill Gurley (board member) and get his thoughts.

Philip's vision for SL is very egalitarian, very net-neutral. I'm not sure how fond he would be of ANY preferential access.
.

I had an analogy for this a long time ago: Imagine a restaurant where anyone can come in and eat for free. If you pay, you can be seated at a table - that's the only difference.

Pretty soon everybody and his dog has heard of this restaurant and has jammed inside to eat for free.

There are so many people, sometimes you can't even get TO your table. Service is laughable, and you pay, but may or may not get your food.

Just how long is that going to last? Well, probably until the owners finally run out of money providing free (if unpredictable) food to the masses, because the ones who have been subsidizing it all by paying for tables have left for lack of service.

From observing the past experiences I tend to think these types of failures (transactions, tp failures) are more to do with concurrency, failure of outside services that LL depend on, and/or hardware failures. I dont think that any of these are SL server software problems as such. So I find it hard to think how they might manage to keep these things to a Wednesday.

Unless you mean the Havok4 rollout? I would hope that has nothing to do with transactions and teleports. I guess it is possible that this rolling upgrade is stressing the servers and impairing their normal services but that should end when the upgrade is complete. I guess if they could have kept that just to a single day - they would have.

Didn't the current raft of problems start before the rollout though?

Contingency Measures to Ensure Service as Second Life Grows
Friday, February 16th, 2007 at 3:04 PM by: Robin Linden

Since September concurrency rates have tripled, to a peak last week of over 34,000. While we love that so many people are enjoying Second Life, there have been some challenging moments in keeping up with the growth, resulting in the now somewhat infamous message “heavy load on the database”. When this happens it usually means that the demand for transmission of data between servers is outstripping the ability of the network to support it.

When the Grid is under stress, resulting in content loss and a generally poor experience, we would like to have an option less disruptive than bringing the whole Grid down. So we’ve developed a contingency plan to manage log-ins to the Grid when, in our judgment, the risk of content loss begins to outweigh the value of higher concurrency. Looking at the concurrency levels, it’s clear heaviest use is on the weekends.

When you open your log-in screen and see in the upper right hand corner Grid Status: Restricted, you’ll know that only those Second Life Residents who have transacted with Linden Lab either by being a premium account holder, owning land, or purchasing currency on the LindeX, will be able to log-in. Residents who are in Second Life when this occurs will only be affected if they log-out and want to return before the grid returns to normal status.

At the same time, new account registrations will be closed.

We hope that we won’t have to implement this contingency plan. As you know, we have regularly improved the Grid’s ability to manage higher concurrency rates, and have comfortably doubled our capacity in recent months. If we determine, however, that we do need to limit log-ins, there will be an announcement made in-world and also on the blog.

Scaling the Grid continues to be our highest priority, and over the next several months we’ll be working on the following:
- additional caching systems
- reducing write load on the central database by partitioning or removing data
- addressing critical bottlenecks
- deploying more internal ‘web services’.

Beyond that we are building ambitious plans for re-architecting the Grid, so that in the future we can realize the full potential of the Second Life Grid to support millions of concurrent users.

Until the servers are in place, it is good practice to check the blog for updates, particularly during peak hours of 12:00 pm. - 4:00 p.m.

I quite like an automated version of Colette's suggestion: when things get really bad, identify the highest traffic sims and randomly disconnect some clients; those with payment info on file can just log back in.

Automatically Disconnect everyone who is logging on with more than one account per IP in high traffic areas.

Then you are more likely to get the bots.

Kinda like flushing them ...

That is the last thing they will ever do. About every corporate account will use 1 outside ip number for all employees logging in. And those seem to be the users they want the most Imagine the reaction of IBM if 1 of their employees can login during peak hours...

But those 2 avatars who just sit and do their sitting animations, how on earth can they stress the system?