Jump to content

Deploys for the week of 2011-06-27


Oskar Linden
 Share

You are about to reply to a thread that has been inactive for 4678 days.

Please take a moment to consider if this thread is worth bumping.

Recommended Posts

@WolfBaginski

You know, when things are fouled up we always seems to get the same Linden Lab apologists telling us how "It ain't so bad" and "They're doing their best" and "SL is a complex programme".  Jack, Elite, a few others and sad to say, you.

Let me put this simply.  Linden Lab has had nearly a decade of experience in this "complex programme".  We pay a lot of money...RL money, to partake of this programme, and I, for one, love it.

BUT

When something as glaring as this happens, and it has happened a fair few times, we do NOT appreciate the ponderous process that ensues, with NOTHING being reported to us via the Grid Status Page.  That is what causes the furore, not that there's been a miscalculation or a simple oversight in the effect a piece of software will or might have.  In the ensuing vacuum you cannot be surprised if folk start to shout.  I had begun to wonder whether the Release Blog was actually being monitored, and I try to be positive and give useable info if I do post - I don't just rant.

Once again we are seeing that Linden Lab have failed signally to learn that communication to your customers is paramount.

If LL had at least acknowledged that there was a possible issue with this release, a lot of hand-wringing would have been avoided and LL's reputation would not have been trashed again.  It is not sufficient to say that an announcement was made as soon as possible after all the committees had met ( those of us in the EU know how very effective that process is).

I have little doubt that this issue will be sorted, and that SL will continue to move ..erm...forward, but the bottom line is still...

TALK TO US!!

 

Link to comment
Share on other sites

  • Replies 50
  • Created
  • Last Reply

Top Posters In This Topic

With the latest rolling restarts in (checks clock) minutes at 05:00amSLT/08:00amEDT, I wonder if the restarts would last... It's not in the Public server releases schedule via Google Calendar on the Second Life Grid status page. I hope this fixes the lag issue in homestead regions.

(Wishes me luck)

Link to comment
Share on other sites

You obviously don't read everything I write about the Lab. It is not unknown for me to get a trifle heated. I can use  all the tricks, dramatic irony, metaphor, bathos, puns, parody, litotes and... satire.

I can make a few wild guesses about why this wasn't spotted earlier. And the communication could have been better. But they're getting a fix out, and I've not seen any sign of the fix breaking something else. So I find it hard to get emotional about this. I know enough to know that I have no chance of understanding the details. It's not impossible that the server code is a tangled mess of accumulated crud, which is the very devil to work on. I wonder if they even know enough to write the specification for a Server 2, in the way that they obviously did for Viewer 2.

But this time, they are getting the job done. Which do you want? Grovelling apologies, or working code?

Link to comment
Share on other sites

@Wolf

Yes, OK, I took the inclusion a bit far by tarring you with the apologist brush, so I apologise for that.

Having said that, Yes it seems that the Homestead I live on is now working better, though I maintain my opinion that some (servers) work better than others.  I guess it might depend on what is happening on the other cores of the server that we are sitting on.

And Yes, also it is far better to have good code put out to correct a problem - I never suggested that LL rush to a conclusion - merely that they tell us that they had noticed!

That server code is a tangled mess of accumulated crud is a back-handed compliment at best, since LL composed the crud.

It seems likely that LL is borderline competent when it comes to viewer design, since before long the Firestorm viewer will be a far better instrument  to use in SL than the (admittedly now much improved) Standard Viewer 2.  Why that would make them any less incompetent at designing and implementing SL2 I don't know and that prospect does not fill me with hope.

That Linden Lab was once populated with bright insightful folk is most likely true, but whether it is so equipped now in entirely another matter.  However, that is off topic and I shall clamber down from the parapet and get on with my (second) life.

@Cincia:

SL is NOT a game.  Sassy is correct; in terms of communication and damage control this was an abject failure of a service company that needs to sort its procedures out.

Link to comment
Share on other sites

When people derive an income from a platform then it's not a game to them.  But this isn't about game or no game but purely good change management practice and putting into practice what is industry standard would allow better continuity especially for those for whom it's a platform for commerce.

Link to comment
Share on other sites

The Server Update today did not fix the probelm of erratic Time Dialtion (caused by the Server Update on June 28th) for Gemini Isle.

I mentioned this to Oskar in the Second Life Beta group chat and also re-opened the support ticket I had for this.

Customer sastisfation... ZERO!

 

Link to comment
Share on other sites

My results after the fix have been dismal as well. All of my full sims are still FAR worse than normal operating performance.

Performance with Second Life Server 11.06.29.234329

Virgins Lagoon 11.6 (should be 7 ish ms)

Virgins Bay    11.8 Ms (should be6-8 ms)

Virgins Landing   12.7 ms  (should be 8 ms)

Virgins Cove  11 ms  (should be 6 ms)

Virgins Reef   8.0 ms  (should be 4.5 - 5.0ms)

 

IN ADDITION:  Nine items in Virgins Cove sim that are for sale have had root prim names deleted, which means the scripts in them will not work.

 

 

Link to comment
Share on other sites

TD back to normal on my homestead (main server).

Spare script time stopped boiinging and back to a healthy 16.0.

But a quick post as soon as a "suspected" issue becomes evident REALLY would do so much avoid all this stress and tension.

Thanx anyhow :) 

 

Apart from that could anyone enlighten me as to why vehicle sim crossing often seems to be of such critical importance here ? I'd have thought it to be an extremely minor consideration.

Link to comment
Share on other sites

If you knew how many boats I have sold in the last several years, you may be shocked at how many residents consider sim crossing to be important. However, I'm not sure what this issue has to do with the performance problems the new server has caused throughout sl.

 

 

Link to comment
Share on other sites

I'm going to stick my head above the parapet and say "It ain't fixed, Oskar".  It's better, but it's not back where it was and there's a lot of sims, some of them Full-regions where performance after that second roll is barely improved at all.

For me, I'd say we have about 80% of the sparetime back, but I wonder just how much of this is caused by providing the tools necessary to count and then limit scripts on a sim?  Perhaps the intial throttle was a bit too severe?

You are at 257371.6, 321842.1, 901.8 in Woods of Heaven located at sim2917.agni.lindenlab.com (216.82.0.150:13001)
Second Life Server 11.06.29.234329.

I'm a long way from being convinced that the pain is over, and that is a surprise.  I just wonder how much we aren't being told?

Now some of you might think that I'm just plain ungrateful; after all it is a much better situation most of us find ourselves in now. 

Well I look at it this way:  I pay a substantial amount of RL money to lease  my sim and I expect my lessor to be honest with me.  In the UK that would be backed up by law, but SL is US, and your laws are different.

I have a feeling as I've said before that this is merely the thin end of a very fat wedge.

My personal script use is high, I admit, but, thanks to good script management, our sim was running as sweet as a nut before this Mono2 performance upgrade was introduced, and now it is just about OK. That IS NOT GOOD ENOUGH.

Oskar, Maestro and Kelly, just WHAT is going on?

Link to comment
Share on other sites

Perhaps the higher script times being reported are  the result of the following  "bug fix"

Improved 'Top Scripts' report

Each object now reports an average script time per frame since the object was created or the last 30 minutes, whichever is shorter. This fixes a lot of odd behavior and misinformation in the previous results as well as makes it easier to determine recent lag sources instead of only immediate lag sources.
It is still not expected for this result to exactly match the 'Script ms' reported in the simulator stats summary as these measure different things on different time scales.

I wonder if this "fix " was kept in place  after the roll back??

Perhaps Oskar or Kelly  could  answer this question plz?

Link to comment
Share on other sites

I have about 1,000 scripts rezzed on my Homestead. I take extreme advantage of mono's bytecode sharing and have rescripted  all of my products to use as many common scripts as possible. I eliminated approximately 500 scripts this way even though each mono script is 64k of memory vs the 16k of memory for the LSL scripts they replaced.

Right now, if my region Time Dilation was stable and sitting at 1.0 (which it should be but is not), I would have a script time of 3 ms. I have no qualms switching everything back over to LSL scripts if what is causing this erratic TD problem is related to mono updates. Unfortunately, Homesteads are 3 per cpu core on a server so that even if I switched everything back over to LSL, I'm still sharing a server with 11 other Homesteads that have not.

My rez day is in a couple of days and I'm thinking that maybe I've been around just a little too long. In 4 years there really has been any improvements... none related to Customer Satisfaction, Innovation, or Respect for the Customer anyways. It seems that I've always been at battle with the the incompetencies and mismanagement of LL.

A CSR stopped by the sim today responding to the ticket I submitted. I probably would have never seen a tech out there if it wasn't for Oskar's efforts (I do appreciate the effort Oskar) but the techs comments have not been reassuring of anything so far. Maybe there are "devs" still working on the situation.

I'll spend this weekend switching everything back over to LSL to see what happens but I don't have high hopes and this will probably the last effort I will put towards trying to maintain a presence in SL. 

Link to comment
Share on other sites

Funny thing...

I recompiled all the scripts on my sim to LSL except for 3 servers that absolutely need to be mono and then reset the sim.

That made the Time Dilation relatively stable in the mid 0.90's. I remove the servers and I get high 0.90's. There's an oocasional drop down into the 0.80's but I attribute that to just being a Homestead sim. Script time is running less than 3ms.

Two avatars TP'd in while I was testing and I watched the TD lower a signicant percentage. When they TP'd out, everything went back to normal. I believe some of the 400 scripts they had attached between them were probably mono scripts.

Funny thing!

Link to comment
Share on other sites

Dear LL, this isn't fixed but I have a great idea!

You see, my existence in SL is almost exclusively that as a merchant.  You know? making and selling things.  If I don't make stuff that's not worth buying, then people won't buy, pretty simple equation.

This recent sim update and it's follow up have had a major negative impact on us, your customers.  It affects the shopping experience and that directly impacts me.

But all is not lost, here's my suggestion.  How about if your salary was index linked to Time Dilation.  This would work great.  If code was good and TD =1.00 you get paid full salary.

Can you work out what you'd be getting paid right now? 

Now factor how that feels for us, the customers of LL while we wait for this recent sim code to be fixed instead of rolled back to a week ago.  There isn't a single good change control practice demonstrated over this past week.

Region: Cape Romano
Uptime: 4 days, 5:49
Dilation: 0.78, min=0.08, mean=0.72, max=1.00, std.dev=0.27
FPS: 30.1, min=4.31, mean=31.6, max=44.1, std.dev=11.9
Combined: 72.6%, 1 min=71.6%, 5 min=88.9%, 15 min=89.1%

Link to comment
Share on other sites

You are about to reply to a thread that has been inactive for 4678 days.

Please take a moment to consider if this thread is worth bumping.

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share


×
×
  • Create New...