Down again?

funny like a tapeworm, maybe...::

bet you 5L that the techie at the server place spilled his coffee on the login server Though it seems it might have just been a router hiccup

I feel sorry for the poor geeks who have to deal with this, hehe. Not fun to be them today.

Got to "contacting user server" this time...


Kyushu

You sure it was coffee and not some other.... umm... substance?

yeah definately a router hiccup. just pinged secondlife.com got a response... great the lindens are probably gonna shutdown the grid for 6 hours while they investigate.

This morning's debalce by the time it was through more resembled a major major set of server crashes than any kind of power outage...and the day after an upgrade too. Go figure. Also followed tonight by yet anotehr crash...oh..excuse me.....probably power outage again. C'mon LL..give us a break and credit for some intelligence..there are those of us who ahve been in this industry for some years. Enough we can sit back, watch behaviour adn tell you what's going on just frmo the symptoms. Personally I wouyld ahve a lot more faith in the customer support if we'd get a straight story about what goes on rather than ppl following corporate orders to trweat us like we know nothing.

Yup, kyushu, that's where I am now. Man, is our spamming hurting the connection?

<<ninja edit>> Whoops, no reply from login database.

Damn, at this rate we're going to need an SL drinking game.

Take a sip for every upgrade. Take a sip and hop on the forums when your sim goes down.

Chug the bottle when all of SL goes down and they blame it on a power failure.

If it was a router hiccup then how come so many who don't even live anywhere near each other can't get in??

I'm having tom collins's.

ANd there's the closed for maintainance message. Whee!!

Is it too much to hope for that they will actually tell us the truth this time??




OMG!! I made myself crack up with that one. ROFLMAO

Preacher needs hugz...we were about to go clubbing!!

I hadn't heard about the power failure till now... that can cause all sorts of issues... last time I admin'd a datacenter we had harddrives going down for weeks after a clean power down off backup power, made for a lot of raid rebuilds... Never know what will pop up after life blood is removed from a machine thats been running 24/7 for a couple + years...

Clubbing! Awww. *huggz* Preacher!

Noooo! I need my secondcrack :<

Well if we do lose some drives lets hope at least one of them has the Impeach Bush guy's resources on it.

Might as well hope for lemonade from lemons.

I've worked in data centers too, and I would not assume drives are going to fail just because the power went out. I have seen isolated problems like that before, but generally with old drives.

I'm still curious to know what modern colo faciluty doesn't have a backup generator, so I can make sure to keep my hardware out of it.


Kyushu

Habib's Falafel Hut and Colocation Facility?

I REALLY doubt that LL would place their servers in a colo facility without ANY back-up power. A more likely scenario is that the colo facility has not been upgrading their back-up power capacity at the same rate as LL has been adding new servers. You can have a nice beefy diesel back-up generator (or two or three or twenty) bigger than most cars in the back of your colo facility, but if you lose power and your servers and all the network and misc. electronic equipment in the building need more power than what your generator(s) can provide, then you're still going to have issues.

If they haven't already, it's time for LL to start looking into a service like Akamai so they can spread the load of serving their static content (images/textures/sounds/animation files) geographically around the world and have redundancy of that content many times over. It would take the concentrated load off their current server farm and result in their current servers only getting hit for logins and when new content is first uploaded and perhaps chat, while delegating the function of serving all of the textures/images/sounds/animations/etc. to the the edge servers spread all around the world. It's worked great for Apple, who uses it to serve the content for their website and the iTunes music store...and all the QuickTime movie trailers on the Apple site.

Back in my dot com days, we used Akamai for serving all of the images and Flash files for the online gaming site I worked for and after "Akamaizing" our content, we were able to increase our capacity well over 500%. It's been a few years, but we were completely satisfied with them and they were the only vendor we used that we never had a single problem with (compared to our $25k/month dedicated hosting "experts" that employed only guys with twenty certification abbreviations after their names, most of whom couldn't change a lightbulb if it wasn't in their certification study guides.)

Anywho...Akamai...just what the doctor ordered, IMHO.

There is likely lots and lots of back up power at the colo facility. Last year Sl was brought down by a failure in power due to an erronious fire detection that shut down all power. This is a good thing actually because in case of fire, you don't want to keep feeding electricity to the probable source of the fire. From the descriptions given by the Lindens at the time, the colo has multiple redundancies including flywheel and diesel generators. Of course it does no good when the alarm system pulls the plug.

Oddly enough another website that I use for my business went down at exactly the same time and now I am a rather miffed camper. Oddly enough both LL and this other site that I am referring to are both based in San Francisco. The gremlins are out tonight!

I would say that scenario is still on LL. If they are growing rapidly they need to assure that their facility has adequate power and backup power to sustain that growth.


Kyushu