Linden Lab’s Shame – 20,000 Crabby Citizens For 5 minutes

by Pixeleen Mistral on 29/12/06 at 9:28 pm

Pathfinder Linden backpedals immediately after boasting of new record

by Pixeleen Mistral, National Affairs desk

Asset_server_downFor those of us in-world at the time, there were the usual signs that something was up. I saw a normally male pirate materialize with frankly feminine figure and oddly textured clothes, another resident complained of having all her money and friends disappear, a visitor to Sanchon was unable to teleport home, and I could not pass notecards to anyone. Was Second Life broken again?

Our fears were confirmed when everyone’s favorite blue box of joy appeared on the screen – Pathfinder Linden was telling the metaverse what was already apparent – something was not right. What makes this humdrum day special is a little dab of irony that appeared on the official Linden blog – Pathfinder Linden begging residents’ indulgence for database problems immediately after crowing about reaching 20,000 concurrent users. In Pathfinders words, “…we’re still working on the database issues… it’s unfortunate (ironic?) that this happened just as I posted this announcement to share the fact that be broke the 20k concurrency mark”. No arguments here.

Based on the overwhelmingly negative response to this blog entry, several hundred residents wanted something more than happy talk about the excitement of uncontrolled growth and unplanned grid outages. There were over 300 comments posted in 3 hours as we went to press – here is a small sample of the sentiments expressed

Martha Glitter: “woohoo, 20,000 users, all without bodies, hud attachments, scripts, clothes, WORKING friends list, IMs… yeah, great milestone…. no, maybe we can get this thing to actually become STABLE?????”

Teleute Bolero: “Big woop – the number is there but what about being stable? I was unable to purchase an item nor use my inventory as “an item was pending” …. log off and on… no body, no money, no inventory, no friends list…. ya, good to know 20,000 people are there but none can see me.”

Ravanne Sullivan: “I guess that is why I can’t TP, can’t cross a sim border, can’t rez an item from my inventory, can’t get a good search…… The list goes on. What good is having 20,000 people online if your paying customers are screaming because they can not preform the simpilist task in world?”

Some speculated that an almost immediate drop in the number of metaverse residents online was caused by the Linden Lab staffers closing logins as the struggled to deal with an unruly – or unresponsive – database.

In any case, after a day when the world was mostly offline due to un-planned asset server maintenance issues, the game god’s excitement at hitting a new peak load may have been tempered by the resident’s concerns with the spotty performance of the world.

Asset_server_down
earlier in the day – sick asset server

Grid_down

10 Responses to “Linden Lab’s Shame – 20,000 Crabby Citizens For 5 minutes”

  1. Spankubux

    Dec 30th, 2006

    Wonderfully objective title. That’s some fine journalisming right there.

    A real newspaper would have noted the underlying causes (bad hardware) once the reason for the failures became clear. But you circus clowns never miss a chance to shit out your agenda onto the pages of this jokeless punchline you call a newspaper.

  2. Urizenus

    Dec 30th, 2006

    ‘bad hardware’. Hmmm, do you think that’s all it is? Because I keep getting emails about these little blue pills that can solve that problem.

    We’re making progress here folks!

    Pop a few of those pills, Spankubux, and you’ll be “journalisming” all over your monitor every time you see a Post 6 layout and/or an article by me.

  3. dildo baggins

    Dec 30th, 2006

    Spankubux, LL needs a good slap around the head. Their QOS is a total joke, and their PR hype a total shitstorm of lies. Sure, shit happens, hardware fails, software has bugs…but with LL it’s a constant stream of this all the while hyping themselves as the future of the 3d web while ignoring their paying customers.

    No company in their right mind would touch this system as a business platform with the reliability they demonstrate…no matter how much their pr flacks and MSM embedded “journalists” and flaky 20 something “Metaverse developers”, “Evangelists” and “Futurists [wtf is that anyway - some wanker who comes up with ideas that someone else has to implement?" persist in trying to suck up to LL by writing constant fluff pieces on it.

    The fact that developers manage to sell this platform to clients indicates:
    1. no risk analysis on the clients part - indicating gross stupidity. The free media roi wave from first mover advantage [American Apparal were smart] has come and gone, never to return.
    2. the developers … ah, well, how can you call yourself a professional when you lie to clients about the true state of the platform you are pitching?

    Sure, POC demonstrates the potential and the 3d sns/web will happen eventually, but this particular platform is a buggy mess of hacked together shite.

    imo it’s not entirely ll’s fault per se. They were doing just fine up until earlier this year..small user base, growing slowly, performance vs stability ok. Then the Vulture Capitalists probably played their card to force the company to go into user acquisition phase to scale it up so they could do the quick flip, without bothering to listen to anyone who happened to say “um, sorry, we are NOT ready for that growth curve yet”.

    Blowback, I think it’s called.

  4. Prokofy Neva

    Dec 30th, 2006

    I wasn’t crabby. What’s not to like? My land is renting, Frogg Marlowe played my favourite song tonight, and Pixeleen slow-danced with me. Ah, Second Life! So it’s a little crashy, meh, the spice to the tamale.

  5. Inigo Chamerberlin

    Dec 30th, 2006

    Fuck you Jack, *I’m* alright?

  6. Alex Fitzsimmons

    Dec 30th, 2006

    To the extent that a tamal (not “tamale”) has a spice that knocks you out for half a day, that’s absolutely correct.

  7. Prokofy Neva

    Dec 30th, 2006

    Oh, well I only lost two houses I spent hours setting out over and over because that rez-faux script is busted, couldn’t save a single notecard or script for days, and in recent weeks, had **every single landmark on thousands of rental box cards*** carefully and individually placed over 2 years (!) *disappear forever* AND have every bit of work renaming, saving, putting dozens of ads in SEARCH etc all get undone AND had all the terraforming get messed up and the autoreturn unset on a dozen sims — well, sure, but then, other than that, Mrs. Lincoln, did you enjoy the play? *Of course* you did, it’s a great play!

  8. Cherokee Jack

    Dec 30th, 2006

    “begging residents’ indulgence for database problems immediately after crowing about reaching 20,000″

    I didn’t see any begging or crowing in that blog post. I saw a simple fact posted about reaching 20k, with no “omg yay we rock!” at all. Just sharing a fact with us. And when things got bad, he simply apologized, warned people what to expect, and kept people updated on the fact that Linden Labs was working on things. And based on the time stamps and my own experience in world, the badness lasted only about 1 hour.

    Sure, the database suckiness was a pain. Sure they need to fix it. But the blog post about 20k and particularly the one before it about the asset server failure details (which had mostly positive comments) were full of useful information that we all should know about.

    Why do you use pejorative and inaccurate terms like begging and crowing? Are you more interested in muckraking than facts?

  9. Mako Mabellon

    Dec 30th, 2006

    ‘bad hardware’. Hmmm, do you think that’s all it is? Because I keep getting emails about these little blue pills that can solve that problem.

    The hardware failure in question (multiple drives failing at the same time) was the type of thing that manufacturers and users of redundant storage hardware fear. The fact that it happened during a restripe, when the system was more vulnerable than usual due to the large amounts of data being rearranged, really didn’t help.

    To be honest, I think Linden Labs handled it in the best way they could – as testified to by the fact that they managed to get it back up with minimal data loss. I suspect much better-known and larger businesses have run into similar problems (you just don’t get to hear about them…)

  10. like_ummm

    Dec 30th, 2006

    Could someone at SLH please rake some muck on these ‘high journalistic values’ nerds? Surely they are working for either the Avatard or LL?

Leave a Reply