[DONE] Rolling Restart for Patched 1.20 Server Deploy, April 17-19
Thursday, April 17th, 2008 at 8:00 PM by: Prospero Linden[Update 2008-04-19 08:30] The rolling restart that will deploy 1.20.1 to the second half of the grid is finished.
[Update 2008-04-18 11:10] The rolling restart to the first half of the grid is complete. Half of the regions are on server version 1.20.1; half are still on the previous server version, 1.20.0. The server deploy will be completed tomorrow on the schedule below.
[Update 2008-04-17 22:30] We have deployed server version 1.20.1 to 1827 regions.
After a few unsuccessful attempts to deploy server version 1.21, we’ve decided to split this into two separate deploys. First, we will deploy a patched version of Server version 1.20. This will include security and performance fixes, including the much-desired reduction memory usage by the physics engine in openspace regions. The plan for this deploy is:
- Thursday 04/17, starting very soon : deploy the 1.20 patches to about 100 regions. Assuming all is well, deploy the 1.20 patches to about 1500 regions.
- Friday morning 04/18, 5-9AM: deploy 1.20 patches to half of Second Life
- Saturday morning 04/18, 5-9AM: deploy 1.20 patches to the rest of Second Life
We are dividing the main deploy into two halves as a result of a variety of constraints. If we do a rolling restart beginning in the morning, we need to stop it well before peak concurrency to avoid overloading the database with additional load provided by the rolling restart. Thus, the Friday morning rolling restart will only cover half the grid so we will have a wide safety margin before the routine load on the database gets too high.
Next week, we will deploy server version 1.21 to the grid. This includes changes in the way regions are assigned to hosts, changes that will allow us to reduce the stress on the central database. More information about the scheduling of that deploy will be posted as it becomes available (probably not until next Monday).
Note: all the versions referenced in this post are server versions, not viewer versions. You will not need to update your Second Life viewer application as a result of this rolling restart.
Update : From Sidewinder at this post, we have a list of the changes in these patches:
Note that items starting with “DEV” were entered internally, and items starting with “SVC” were reported on the public jira.
DEV-13002: Improved physical representation of some thin shapes (As originally reported, this appeared to be a problem with a prim floor becoming phantom after conversion. It turned out to an inaccurate physical representation that left “holes” in the floor in some areas)
SVC-1531: Physical representation of dimpled, hollowed spheres improved
DEV-12725: Camera position is no longer affected by phantom objects or avatars (resolution of this bug removed a cause of rapid shifts in camera position for reasons not obvious from the scene in view)
DEV-12841: Removed a simulator crash mode
DEV-13080: Reduced tendency of avatar’s feet to fold up to waist climbing stairs
SVC-1989: Avatar no longer becomes phantom when sitting
Improved performance of openspace regions
April 17th, 2008 at 8:13 PM
Sounds good. Myself, I can handle a rolling restart if it means patches. Much improvement over the old-days 1/2 day offline to do so. No complaints.
April 17th, 2008 at 8:27 PM
Sounds good to me too. It’s what some of us in openspace sims asked for…a special rollout that address’s our problems, without having to wait another week for the .21 stuff to be fixed….so you shouldn’t hear much backtalk from us…..
April 17th, 2008 at 8:31 PM
Thanks for the info!
April 17th, 2008 at 8:31 PM
Well I suport this ………………good luck! and Thank You
April 17th, 2008 at 8:41 PM
I hope this agony results in a success and we can call a halt to any changes for at least a month.
What transaction manager are you using for L$ transactions to ensure recovery on failure? BEA Tuxedo®? BEA WebLogic®? Surely your not trying to home roll something this critical. And you must be doing something because such a system as SL has a solid commercial transaction manager platform as a critical core requirement.
April 17th, 2008 at 8:51 PM
Great, glad to hear the problems will be fixed. I downloaded the 1.20 but had to go back to 1.19.1 so good luck. I would like to use 1.20. Also thank you for taking care of the openspace reqests also.
April 17th, 2008 at 8:56 PM
Keep up the fantastic job LL! Second Life is improving all the time and I really appreciate the amount of work you guys do to make it a fun and enjoyable experience for all of us
I LOVE SECOND LIFE!!!!!
April 17th, 2008 at 9:08 PM
@5 Ann:
“call a halt to any changes for at least a month”
I have bad news for you… The Havok4 team is, as promised, continuing to work on open bugs, and already has another version on deck that resolves twelve or thirteen *more* physics-related issues (not sure if they’ll all pass QA testing, so the exact number may float a little by the time we deploy it). We have every intent of releasing these changes as soon as we can, to resolve as many open issues as we can, which will hopefully be next week
Best regards,
Sidewinder
April 17th, 2008 at 9:10 PM
26 Casandra: This is a server-side update. The simulators run at the Linden Lab data centers and are not the same part of the system as the viewer software that you download to your computer. I do understand that this can be somewhat confusing since the version numbers are often pretty similar on the simulator (server) and viewer (desktop or notebook computer) side of the system. /Sidewinder
April 17th, 2008 at 9:22 PM
“SVC-1989: Avatar no longer becomes phantom when sitting”
That sounds really, really bad. Maybe I’m misunderstanding how this worked, but could that not destroy the functionality of sitting as a poor man’s shield?
April 17th, 2008 at 9:25 PM
Does this upgrade have anything to do with my inability to teleport to certain places, and my group list showing as ‘none’ yet the rolls for my groups, still in my profile, show both me as a member and the option to join?
April 17th, 2008 at 9:48 PM
Question:
Are you EVER going to solve the “groups not loaded = failed teleports” issue?
75% of my teleports and SIM-crossing end in me being booted, and that’s in a good day! The situation has gone to a stage where I dread (yes! DREAD!) teleporting or SIM-crossing…
It is tiresome to keep on trying and being booted over and over again for at least half and hour (if I’m lucky) so for me to just be able to teleport somewhere to chat with friends… What’s the use of a community if you’re being held HOSTAGE in a single SIM???
April 17th, 2008 at 9:54 PM
The rumor on the virtual street is that Anonymous is now targeting the asset servers through some sort of inworld scripting exploit.
Anyone in the know care to offer a comment or clarification on this? It is possible? Probable? Could the Second Life experience be worse then it is already?
There has to be some excuse for the daily asset server problems. Surely Linden Labs hasn’t made such a mess of things without outside help.
April 17th, 2008 at 10:00 PM
BTW, just wondering, is it possible to send people back home or to where they were INSTEAD of kicking them out when a failed teleport occurs? I’m really annoyed by the whole “failed teleport - clicking close - restarting SL - wait for all to load” thing…
And sometimes I get this message telling me I can’t go where I’m trying to login and that I should teleport home… The irony is that it ALWAYS happens when I trying to login straight to my Home… !!!!
Ò.ô
April 17th, 2008 at 10:01 PM
@13 Ric : I haven’t heard that rumor. To my knowledge, there is no targeting of the asset server nor any asset server scripting exploits. (I will add the caveat that I don’t know everything that’s going on at the Lab; there are too many Lindens for any one of us to know that right now.) There _is_ a scripting exploit that can crash sims that is being addressed by the 1.20.1 server being rolled out right now.
The daily problems we have aren’t asset server problem so much as central database problems. When the central database is having issues, everything else will tend to struggle along behind it. As I’ve mentioned above and elsewhere, we have a number of things i progress to alleviate load on the central database. One of the biggest sources of gratuitous load right now is our spare machines asking the database over and over again for regions to run. Even though they do this only every few minutes, it adds up. The server version 1.21 includes a new way of assigning hosts to regions that will eliminate this load to the database. Meanwhile, we have new and beefier database hardware being shipped to us, which will buy us some more time. Longer term, the architecture that you can learn about in the AWG office hours hosted by Zero Linden (pieces and early forms of which are already going in to the production grid) will further reduce our dependence on a single central database.
April 17th, 2008 at 10:02 PM
Michael : which client are you using? The vast majority of failed teleports I’ve seen involve me being stuck back at where I’ve started. Sometimes it fails partway through and as things are right now you get kicked out, but that’s the exception rather than the rule in my experience.
April 17th, 2008 at 10:05 PM
@13
Its highly probably Ric, there are alot of jealous people out there and because they feel so small that they are not able to make something out of their lives, they feel that they have to ruin everyone elses lives to make themselves feel a bit more in control. There are people who fail and pick themselves up from the dirt, and there are those who fail and keep falling with everything they do… those are the real losers in life.
April 17th, 2008 at 10:06 PM
If it was any /b/ tard or other such group they wouldn’t be quiet about it. And also simply put this has been building for a long time. I’lll post the list of the last few months that Guru Dreamscape first posted.
Second Life In-World Service Issues
Thursday, April 17th, 2008 at 9:31 AM by: rheyalinden
[Resolved] In-world Issues
Wednesday, April 16th, 2008 at 7:50 PM by: rikalinden
[Completed] Services and Logins will be down for ~30 minutes
Tuesday, April 15th, 2008 at 11:49 AM by: Kate Linden
[Resolved] Temporary interruption of services
Monday, April 14th, 2008 at 5:03 PM by: Kate Linden
[RESOLVED] Database struggles, interruption of several inworld services
Sunday, April 13th, 2008 at 8:20 AM by: Lotte Linden
[RESOLVED] More Database Unpleasantness
Friday, April 11th, 2008 at 11:43 PM by: Twilight Linden
[RESOLVED] Failed Teleports and Region Crossings
Friday, April 11th, 2008 at 11:17 PM by: Twilight Linden
[BUMPED] Object to object emails failing in-world
Friday, April 11th, 2008 at 6:54 PM by: Matthew Linden
[CLOSED] Re-enabling Services
Friday, April 11th, 2008 at 3:04 PM by: Teeple Linden
[CLOSED] Login issues and some In World Features not available.
Friday, April 11th, 2008 at 7:14 AM by: Chiyo Linden
[RESOLVED] Two Gremlins Bite the Dust
Thursday, April 10th, 2008 at 11:00 PM by: Teeple Linden
[RESOLVED] ~700 Regions Offline
Thursday, April 10th, 2008 at 11:43 AM by: Twilight Linden
RESOLVED - Upgrading Central Database System Wednesday, April 9th 7:00 a.m. - 8:00 a.m.
Wednesday, April 9th, 2008 at 7:00 AM by: Kate Linden
**This didn’t seem to help much**
[RESOLVED] Regions Down
Tuesday, April 8th, 2008 at 7:41 AM by: sejonglinden
[Resolved] Grey Profiles Reported
Monday, April 7th, 2008 at 5:21 PM by: Kate Linden
[RESOLVED] Logins & Transactions Working Intermittently/Peak Hours Info
Monday, April 7th, 2008 at 2:44 PM by: Kate Linden
(Completed) Profile and Group Services Temporarily Unavailable
Monday, April 7th, 2008 at 1:05 PM by: Teeple Linden
[Completed] Profile and Group Temporary Changes 1:00 PM PST
Sunday, April 6th, 2008 at 9:40 AM by: Kate Linden
[RESOLVED] Intermittent inventory and login issues
Sunday, April 6th, 2008 at 3:40 AM by: Lotte Linden
[RESOLVED] Intermittent Region Issues
Sunday, April 6th, 2008 at 12:06 AM by: Elle Linden
[RESOLVED] Second Life is currently down - logins are disabled
Saturday, April 5th, 2008 at 12:55 PM by: Teeple Linden
[MERGED] Various Inworld Issues
Saturday, April 5th, 2008 at 5:25 AM by: sejonglinden
Logins Disabled until Further Notice
Friday, April 4th, 2008 at 10:22 PM by: Teeple Linden
[UPDATED] Various Inworld Issues
Friday, April 4th, 2008 at 7:33 PM by: sejonglinden
[Resolved] Various Issues
Wednesday, April 2nd, 2008 at 12:03 PM by: Joppa Linden
[Resolved] Various Inworld Issues
Tuesday, April 1st, 2008 at 12:05 PM by: rikalinden
[ALL-CLEAR] Second Life Phone Support Issues
Tuesday, April 1st, 2008 at 10:51 AM by: rikalinden
[Resolved] Second Life In-World Service Issues
Monday, March 31st, 2008 at 2:33 PM by: rheyalinden
[Resolved] Live Chat Temporarily Unavailable
Monday, March 31st, 2008 at 1:32 PM by: rheyalinden
[Resolved] In World Services Down/Website and Support
Sunday, March 30th, 2008 at 2:05 PM by: Kate Linden
[All Clear] Transactions Timing Out
Sunday, March 30th, 2008 at 9:39 AM by: Kate Linden
[Resolved] Support Phone Lines Unavailable
Sunday, March 30th, 2008 at 9:23 AM by: Kate Linden
[RESOLVED] Inworld Money Transactions Problematic at the Moment
Saturday, March 29th, 2008 at 2:30 PM by: Chiyo Linden
[RESOLVED] Asset Server Issues
Friday, March 28th, 2008 at 3:45 PM by: Elle Linden
[All Clear] Inworld Issue Affecting Transactions
Thursday, March 27th, 2008 at 1:13 PM by: Twilight Linden
[ALL CLEAR] In World Service Disruptions Under Investigation
Tuesday, March 25th, 2008 at 11:37 AM by: Teeple Linden
[RESOLVED] Disruption in inworld services
Monday, March 24th, 2008 at 9:30 AM by: Lotte Linden
[RESOLVED] Loading Issues
Saturday, March 22nd, 2008 at 12:16 PM by: Joppa Linden
Some Regions Currently Unavailable
Saturday, March 22nd, 2008 at 5:37 AM by: Chiyo Linden
[RESOLVED] Multiple Known Issues
Friday, March 21st, 2008 at 12:49 PM by: Joppa Linden
Region Data Corruption / Loss Problem
Wednesday, March 19th, 2008 at 5:49 PM by: Kate Linden
I see gray people… and they walk around like they don’t know they are gray.
Wednesday, March 19th, 2008 at 3:00 PM by: Chiyo Linden
[RESOLVED] - Some Regions Not Reachable
Wednesday, March 19th, 2008 at 12:27 AM by: Matthew Linden
[Resolved]Unable to Login
Tuesday, March 18th, 2008 at 4:28 PM by: Kate Linden
[STABILIZED] Reports of Stale Transactions
Monday, March 17th, 2008 at 7:22 PM by: Kate Linden
[Resolved] Inworld Services Interrupted
Monday, March 17th, 2008 at 2:30 PM by: Kate Linden
[Resolved] Login problems
Monday, March 17th, 2008 at 7:05 AM by: Chiyo Linden
[Resolved] Interrupted In World Services
Thursday, March 13th, 2008 at 4:14 PM by: Kate Linden
[RESOLVED] Account Detail Reports Lagging by Roughly Five Hours
Monday, March 10th, 2008 at 10:13 AM by: Teeple Linden
[RESOLVED] A number of regions currently unreachable
Monday, March 10th, 2008 at 7:00 AM by: Matthew Linden
[RESOLVED] Events page not allowing residents to add new events
Monday, March 10th, 2008 at 2:49 AM by: Matthew Linden
[Resolved] In world services interrupted
Sunday, March 9th, 2008 at 11:44 AM by: Kate Linden
[Resolved] Service Interruptions -L$ Transactions
Sunday, March 2nd, 2008 at 3:30 PM by: rikalinden
[Resolved] Support portal unavailable
Saturday, March 1st, 2008 at 3:47 PM by: Kate Linden
[RESOLVED] Database issues - Transaction Histories not available.
Friday, February 29th, 2008 at 10:52 AM by: Joppa Linden
[Resolved] Billing Issues
Thursday, February 28th, 2008 at 8:41 AM by: Data Linden
[Resolved] In-world issues being addressed.
Wednesday, February 27th, 2008 at 3:22 PM by: Joppa Linden
[Resolved] Region Data Corruption / Loss Problem
Saturday, February 23rd, 2008 at 2:10 PM by: Lotte Linden
[ALL CLEAR] Rollback on Multiple Regions at 7pm PST Tonight
Wednesday, February 20th, 2008 at 8:40 PM by: Teeple Linden
[Resolved] Network Issues Affecting ~450 regions
Wednesday, February 20th, 2008 at 6:59 PM by: Joppa Linden
[Resolved] General network related issues
Monday, February 18th, 2008 at 5:36 PM by: Kate Linden
[RESOLVED] Inworld Services slowdown
Monday, February 18th, 2008 at 1:28 PM by: Lotte Linden
[RESOLVED] More DNS Woes
Tuesday, February 12th, 2008 at 5:05 PM by: Teeple Linden
[Resolved] Support Unavailable
Tuesday, February 12th, 2008 at 4:15 PM by: Kate Linden
Intermittent Issues with In World Voice
Tuesday, February 12th, 2008 at 3:09 PM by: Teeple Linden
[CLEARED] Some Residents Experiencing DNS Failures on Login
Tuesday, February 12th, 2008 at 10:31 AM by: Teeple Linden
[Resolved] Logins Unavailable/TP
Sunday, February 10th, 2008 at 2:00 PM by: Kate Linden
[RESOLVED] Intermittent issues within Second Life and our website
Saturday, February 9th, 2008 at 10:35 PM by: Frontier Linden
[Resolved] Support Portal Outage
Saturday, February 9th, 2008 at 5:27 PM by: Elle Linden
[Resolved] Logins Unavailable
Thursday, February 7th, 2008 at 5:06 PM by: Kate Linden
[RESOLVED] More Intermittent Website Problems
Thursday, February 7th, 2008 at 1:51 PM by: Teeple Linden
[Resolved] Inventory Issues : Top Level folders from Library
Wednesday, February 6th, 2008 at 2:41 PM by: Joppa Linden
[RESOLVED] Stipends and group payouts late
Tuesday, February 5th, 2008 at 10:29 AM by: Lotte Linden
Region Data Corruption / Loss Problem Resolved
Monday, February 4th, 2008 at 10:19 AM by: Hamilton Linden
[RESOLVED] Land search and land sales not up to date - Rolling restart in progress.
Friday, February 1st, 2008 at 6:30 PM by: Matthew Linden
Map Search or Image May Not Be Available
Thursday, January 31st, 2008 at 10:40 PM by: Kate Linden
[Resolved] Inventory Database Unavailable For Some Residents
Thursday, January 31st, 2008 at 4:39 PM by: Kate Linden
Viewing Friends Online on Website Temporarily Disabled
Wednesday, January 30th, 2008 at 8:09 AM by: rheyalinden
[RESOLVED] Support Portal Unavailable
Tuesday, January 29th, 2008 at 9:04 PM by: Kate Linden
[RESOLVED] Intermittent website failures
Sunday, January 27th, 2008 at 6:29 AM by: Lotte Linden
[RESOLVED] Difficulties logging in to Second Life and connecting to secondlife.com
Thursday, January 24th, 2008 at 2:54 AM by: Matthew Linden
[Resolved] Logins
Tuesday, January 8th, 2008 at 6:25 PM by: Kate Linden
This is all from Linden Lab and we can’t put the blaim on someone else
April 17th, 2008 at 10:23 PM
But what about fixing the ||TargetOmega script ???
April 17th, 2008 at 10:25 PM
As far as I’m aware, there is no large-scale /i/nvasion from /b/ or the Anonymous collective. As always, there may be small groups from /b/ performing unofficial raids and assaults against the grid through whatever vector they find. However, insofar as the /b/ name applies, they’re just griefers. Could just as easily be Goons or any other web-savvy group or individual.
In unrelated news… LL, keep up the good work, keep those patches rolling… and once your two major projects (havok4/mono) are at a stable level, dedicate a project entirely to the central database. I know it’s boring, I’m a CS major who hopes to join you someday, but that thing is the cornerstone of your problems right now, with the asset server not far behind it. Don’t let newer projects (like Dazzle) get in the way of fixing your core functionality.
Prospero, Sidewinder, and all other ‘commenting’ Lindens, keep up the feedback - it’s what we need right now, as people and not just users.
April 17th, 2008 at 10:26 PM
@19 Netsah: We have what we believe is a fix for most of the llTargetOmega issues in test now. The Havok4 physics team has a build with over a dozen fixes in it that is in the test process, and we hope to have that deployed next week.
The difficulty of reproducing that problem on demand makes it so that we cannot be sure it will fix all cases, but we believe it should make a significant improvement, and based on the results we will see if there are other cases we need to work on.
/Sidewinder
April 17th, 2008 at 10:44 PM
bought lindens but didnt get them..its on my account but didnt recieve them. If I see the bank transactions tommorrow in RL and dont have them in my sl account tommorrow then thats a red flag I hope that isnt another blooper and at one point i couldnt report so if you can report what do you do? Serious…..
thansk you :((
April 17th, 2008 at 10:47 PM
Sidewinder or Prospero, can you REACT on point 18 ? or is this to factual for LINDENLABS ?
cosa
April 17th, 2008 at 10:51 PM
Will this address the bug that appeared a few weeks back where for unknown (to me anyway) reasons, an open platform skybox seems to shift position every 30 seconds or so making an avatar move (while standing) to compensate?
April 17th, 2008 at 11:27 PM
so is 85162 this new version?
April 17th, 2008 at 11:32 PM
What version number is the new server update?
April 17th, 2008 at 11:43 PM
Will all these new improvements fix the disappearing skirt bug? I’m sure merchants have been loosing money over this since those of us who wear skirts no longer dare to purchase them. I know Pastrami Linden said it was due to “package” loss and we should be on a high end wi-fi connection or broadband connection. I have a pc and am on the highest end broadband connection that can be bought and I still have disappearing skirts! Please do not say the issue is closed because I know I am not the only one out here with this problem. Yes a jira was filed. It was “closed” without being assigned!
April 17th, 2008 at 11:43 PM
So it is going to be “weeks” before the grid is “possibly” stabilized?
April 17th, 2008 at 11:50 PM
“DEV-13002: Improved physical representation of some thin shapes (As originally reported, this appeared to be a problem with a prim floor becoming phantom after conversion. It turned out to an inaccurate physical representation that left “holes” in the floor in some areas)
SVC-1531: Physical representation of dimpled, hollowed spheres improved
DEV-12725: Camera position is no longer affected by phantom objects or avatars (resolution of this bug removed a cause of rapid shifts in camera position for reasons not obvious from the scene in view)
DEV-12841: Removed a simulator crash mode
DEV-13080: Reduced tendency of avatar’s feet to fold up to waist climbing stairs
SVC-1989: Avatar no longer becomes phantom when sitting
Improved performance of openspace regions”
Glad to hear it.
April 18th, 2008 at 12:06 AM
OFFTOPIC:
Second Life Viewer (updated) never shows the login/info screen but keeps using the CPU at 40-80% forever.
I cannot login into the forum because the “Resident login” page doesn’t show up (”connection failed”).
Even SLeek can not log in: “The request was aborted: The request was cancelled”
April 18th, 2008 at 12:14 AM
RE:#25 Well the sims were 1.20.0.83892 so if you are seeing 1.20.85162 then I would surmise that is the new version number thanks for posting it as well as I couldnt find atm a updated one.
April 18th, 2008 at 12:28 AM
I seem to have problems with voice in the 1.20 I don’t even have the option to activate it, it says Voice chat not available.
Every log in means I have reset all prefs they do not save. Even just where to save chat logs has to be reset.
Crash rates for me are very high also and I quite often end up with the viewer locked on my screen forcing a restart on my PC.
So I wish you luck with the patches because if it doesn’t help these issues I shall have to go back to the regular veiwer.
CPU: Intel Pentium 4 (Unknown model) (2994 MHz)
Memory: 2046 MB
OS Version: Microsoft Windows Vista (Build 6000)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: GeForce 7900 GS/PCI/SSE2
OpenGL Version: 2.1.1
April 18th, 2008 at 12:43 AM
I know this is a good thing and we should all be weloming and thankful.
But, right at the start of the weekend! I hope and pray it doesnt spell a weekend of strife and chaos ahead, thats becoming far to much the norm. I know you guys are working so hard to fix and maintain SL for us but I’m also sure you can understand and sympathise why so many *myself included winge moan and get the voodoo dolls out at times.
Here’s to hoping that it all runs smooth and those lucky openspace peeps get their wishes, I just wish 4gig of ram were enough to sate SLs hunger and my skirt would stop vanishing, and alpha textures worked and and and and so forth.
X R
April 18th, 2008 at 12:59 AM
It’s about a month SL is going very bad. Ok SL is a very complex and innovative platform but I think Linden Lab need to start to understand SL is not a game. We put money on it (personally I have invested about 4000 US$ in the last 2 months). So it’s clear Linden Labs is selling us “professional services” but in which way?
Professional services have “services level agreements”, redundant connectivity (remember the last week end SL offline for a day?”, staging servers and so on.
I’m wondering why I’m paying for fees and advertises for every minutes of the day when SL is working no good.
Linden Lab takes care of their money so much but what about our transactions failed and the hours offline. What about my trouble tickets opened 10 days ago still without any answers?
Now we will have at least another bad 10 days in-world and we all just know they won’t be the last. This is not the way to run this business.
April 18th, 2008 at 1:27 AM
It’s over. Everyone cash out and let’s move on.
April 18th, 2008 at 1:45 AM
I have suffered certain expensive objects,vehicles (apolon motors helicopters/planes have reportedly grid wide stop having abilities to fly up in certain sims and before all that his vehicles were top of the line so its not his fault,and autoretun results in object loss instead of back to the lost and found folder)in sl to stop working due to changes in sl so that would be the only compaint i would have but if it means all this leads to a better and better experience using the viewer etc etc,then you guys have my support
I’m pretty excited to see in like 2 years from now or so where sl will be…i hope it will be spectacular!:)
April 18th, 2008 at 1:56 AM
Thank you for pushing out this version with the open space performance problem fixed
April 18th, 2008 at 1:56 AM
I think they are doing the best job they can and I believe that within a couple of years SL will be the most realistic and amazing virtual world that you have ever seen, it will go down in history as one of the major transformations in the evolution of mankind and our culture. Everyone will be on Second Life in the future - It will be THE virtual 3-D interactive platform.
We should continue to support Linden Labs through the hard times and cheer them on to keep up their great work! Lets not forget how ground breaking SL is and don’t let these small obstacles in their progress undermine how big an undertaking this is.
*stands up and raises a glass to Linden Labs*
April 18th, 2008 at 1:58 AM
@23 cosa nostra
What kind of reply do you expect? To sum up the headings and point at problems that are very obvious, to all, isn’t doing anything interesting.
Prospero has, several times, talked about the problems and solutions, even in this thread, so, to me, they have already reacted.
With an as active userbase as the one in SL, problems won’t go by unnoticed. And to me it seems obvious that LL employees read the blogs and react to them in ways that they can.
Even if I’m just guessing a little, I’d say it’s not too unlikely that the LL employees have gotten strict instructions regarding what to write and how to write it, making the comments into what they are.
To get a little on topic, I can only say that I’m looking forward to the upgrade and the future deployment of the new database architecture. I’ve done some work regarding database concurrency myself, and it’s easy to run into problems with a much smaller load than the one in SL if the design isn’t close to optimal for the application.
April 18th, 2008 at 2:07 AM
Will Siggy Romulus’ Swimmer ever be fixed? I always felt if filled an unexplicable void in the default avatar behavior, and I still find it very hard to believe that AVs can wink (and wank) but, once in the water, they behave like boiled potatoes.
April 18th, 2008 at 2:09 AM
thanks for the in world warning… what hapened to that? Just logged me right off saying sim going down.
April 18th, 2008 at 2:15 AM
RE: #32 Dan this is a server update not a client side viewer update. Your references about you having client problems and guessing you are using the RC since you said “I shall have to go back to the regular veiwer.” Is not what this notice was about this has to do with the sim server code update not what you download and install on the user end. and quote…
Sidewinder Linden Post #9: This is a server-side update. The simulators run at the Linden Lab data centers and are not the same part of the system as the viewer software that you download to your computer. I do understand that this can be somewhat confusing since the version numbers are often pretty similar on the simulator (server) and viewer (desktop or notebook computer) side of the system. /Sidewinder
April 18th, 2008 at 2:32 AM
mmmmmmmm
I’m having really bad issues now (with the sever side) and I’m using the
./secondlife –channel “skipRC1″ command to run SL as well.
April 18th, 2008 at 3:04 AM
Hi there, since i install Second Life 1.20.1 (84760) Apr 15 2008 12:13:49 (Second Life Release Candidate), my sl crash using antialising, crash so badly after i login i have a imidiate crash, till i turn anti… off.
Here is my machine all drivers up to date!
CPU: Intel Core 2(quad) Series Processor (2400 MHz)
Memory: 3583 MB
OS Version: Microsoft Windows XP Service Pack 2 (Build 2600)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: GeForce 9600 GT/PCI/SSE2
OpenGL Version: 2.1.2
Don´t know if its a client issue or something you can fix in server updates! Keep up the good job Lindens!
To all , Have a nice day.
April 18th, 2008 at 3:21 AM
Oh BTW on the skirt issue, I have found that the old rebake *ctl+alt+R* seems to work if you persist at it, or swapping skirts back and forth, though they can just suddenly vanish anytime ut seems,
Well on my screen its sorted this way, I just have to wear grannies old bloomers under just incase though.
Oh and I appreciate the fact Linden Labs have left comments open on some of the more frustrating blogs too, whether accidental or not , it’s better to whinge and spit vitriol where it should be directed and not at poor old Torley.
*thinks I must be unwell, 2 posts and no whinging* heads off to the med cabinet!
April 18th, 2008 at 3:59 AM
Avatar phantom on sit is not a glitch?!
April 18th, 2008 at 4:08 AM
@ 46 well thats been happening for a few weeks now……….
April 18th, 2008 at 4:37 AM
I really hope object rotation (llTargetOmega) gets fixed soon. I’m glad to hear the Havok4 team is working on it.
I have noticed that early adopters of havok4 are not having any rotation issues. Just the rest of us. And I mean all of us. If the team is having trouble reproducing the bugs, just go to any sim that did not participate in early tests of havok4 and rez anything that uses llTargetOmega and it will not spin in *any* of the available viewers.
This kinda thing makes it very frustrating for content creators as all our customers hound us (understandably,) at the same time trying to find out why their products are not working. Plus it kills sales for our stuff while the issue continues.
I appreciate the work LL does very much and am not whining, but I wish this particular issue was addressed more clearly in a blog post so I have a place to refer my disgruntled customers.
Also real quick, apparently we have two 04/18’s this month. (I tease because I love I swear.)
# Friday morning 04/18, 5-9AM: deploy 1.20 patches to half of Second Life
# Saturday morning 04/18, 5-9AM: deploy 1.20 patches to the rest of Second Life
April 18th, 2008 at 5:08 AM
Sidewinder or Prospero, can you REACT on point 18 ? or is this to factual for LINDENLABS ?
It is just factual; what reaction do you want? Yes, we know that we’ve had repeated problems with service interruptions in the last several weeks. We are no more happy about this than anybody else. This is why we have various projects in progress to reduce load on the central database, which is the source of most of the problems listed. (Some of the problems listed are from other things, including external causes such as network issues.)
Also, just a note: it’s “Linden Lab”. (No s on the Lab.)
April 18th, 2008 at 5:11 AM
Re; two 04/18’s… that’s just wishful thinking. Time flies too fast. I can’t keep up. I suppose I should update the blog post above to reflect reality. …sigh…
April 18th, 2008 at 5:13 AM
@ 48 “I have noticed that early adopters of havok4 are not having any rotation issues. Just the rest of us.”
That’s weird you know, because my pillow talk raft, which is a multiprim linked object with an llTargetOmega spins nicely in my early adopter region, but looking in the pillowtalk sim and a few others too that were not early adopters I see theirs are stationary.
Anecdotal, but interesting.
April 18th, 2008 at 5:31 AM
@51 I was just basing my assumption that early adopters weren’t having the rotation issue on the fact that two of my friend’s sims that were part of that program are not having rotation issues. I obviously have not checked all of the early adopters.
I do remember though that there was an issue when the early adopter regions first got havok4 that things were not rotating. Until LL found and installed a fix. Feels like whatever research that was done then got lost and was not implemented before the rest of us got the new physics engine.
Regardless, I’m sure it’ll be fixed soon and until then I can refer my customers to the JIRA report on this issue for more information and to Vote/Watch that topic: https://jira.secondlife.com/browse/SVC-1910
April 18th, 2008 at 5:37 AM
#11 & #12 teleports failing https://jira.secondlife.com/browse/VWR-6265 Marked as critical and assigned.
#11 groups not loading https://jira.secondlife.com/browse/VWR-2633 also critical and assigned
#32 preferences not saving https://jira.secondlife.com/browse/VWR-6265 Fix pending
#32 voice not able to be enabled https://jira.secondlife.com/browse/VWR-6286 Workaround
#44 RC crashes https://jira.secondlife.com/browse/VWR-6343 Marked as a showstopper. This happens to me too, so I just go back to the Official Viewer, but, last night, I did find some updated drivers for my Nvidia 8600 so I am going to try again today. I will also post on JIRA
I hope that all the people who are having these problems have or will vote for them on JIRA, adding details about their system, if relevant. Please, we all know that if something affects alot of people then it’s likely to be a bug and not a problem with our personal computer setup.
April 18th, 2008 at 5:37 AM
Prospero can you clarify what the server version should be as I read the blog again its saying 1.20.1 ???
Quote: We have deployed server version 1.20.1 to 1827 regions.
I thought we are staying at 1.20.x.x till the 2nd half of the splitted patch. Im seeing 1.20.1.85162 over at the mature sandbox. According to the next line we are going to be a patched 1.20 still so even though its 8am and 39 hours later without sleep
its even confusing me.
April 18th, 2008 at 5:38 AM
Voice still doesn’ work.
April 18th, 2008 at 5:39 AM
My Voice doesn’ work either…
April 18th, 2008 at 5:44 AM
@Sidewinder and Prospero: Thanks for continuing to work the Havok 4 issues. I hope one of the fixes will address the “sinking while hovering” bug. Also, my teleports have been MUCH improved (no crashes, and no delayed tps) by using Nicholaz Beresford’s ED-e viewer patch. The WindLight team should look into this one!
April 18th, 2008 at 5:44 AM
Well,
I don’t like this attitude at all of LindenLabs. If you want to change something then notify in advance, and then keep your schedule.
Somehow its impossible to predict anything anymore inworld. I am leading a club, and its somehow not possible to do anymore events, when anything changes alltime and alltime accidently. installing new software - it doesnt work - back to the old, then a new patch, then rolling restart again, no predictions about which of 1400 sims would be affected and so on…..
The last two weeks already is a major backlash in inworld economy. At least what i heard from many others.
Can’t you act at all like a normal company? Reliabilty, stability and predictability should be the first thing. Not the last.
I am sorry for the inconvinience …..
April 18th, 2008 at 5:51 AM
improvements look good ( fixes) I look forward to being able to login again soon. God speed team. Woot ZIm looking for happier SL days ( well perhaps not this weekend)
April 18th, 2008 at 6:03 AM
@53 : the version we’ve been trying to roll out is 1.21.0. *This* version, with the patches, is 1.20.1. The previous version is 1.20.0.
It’s a bit confusing with all the 0’s and 1’s….
April 18th, 2008 at 6:40 AM
@18: weekends usually mean highest load and thus highest chance of something being borked. take out the weekends, what stays isnt really enough data to say something reliable, other than problems seem to occur on a somewhat repeating pattern xD
April 18th, 2008 at 6:43 AM
cannot believe its going ahead, things are rezzing so slow for me and a friend right now
April 18th, 2008 at 6:48 AM
Sidewinder….If you want some really good examples of what the llTargetOmega scripts are up to just go to anyone of my stores in world.. I have tons of examples in them showing all different levels of what is going on..some work..some only work when you walk right up on them..some work when you re-rezz the objects and some just flat out wont work at all no matter what you do reset the scripts..re rezz etc…
April 18th, 2008 at 6:54 AM
Those having voice problems in the Release Candidate see VWR-6286 for a possible fix (and vote
April 18th, 2008 at 7:17 AM
I too believe that this could be a huge achievement, but during this painful learning curve which we share with the Lindens, it is costing us money in down time, but making them money because there is no compensation system … and we are talking megabucks here from what I understand.
So come on, how about a really nice gesture Lindens, give us back a percentage of our tier based on the downtime we have experienced.
Yes it would cost you, but think of the goodwill it would generate …
…
and, yes, I have applied for a PR job with you! LOL
April 18th, 2008 at 7:26 AM
@21
Thanks for the answer.
Because you see, like other friends of me inworld, I could see the ||TargetOmega working again on SL 1.19RC, after not seeing it working for months, and now it doesn’t work anymore again for me with the normal viewer 1.19.4 and also not working on 1.20RC…
Thanks for your great work !!!
April 18th, 2008 at 7:29 AM
yay.. i think this is my first on topic post.. before and after the rolling restart, i get:
[7:17] Can’t reposition — permission denied
when i try to move my own (sculpted) prims and they snap back in their old postion! WTF *rubs eyes*
i hope this is know, adressed and fixes asap,
greetings, jaco
April 18th, 2008 at 7:32 AM
omg.. and sorry for spelling ^^
April 18th, 2008 at 7:33 AM
@62 Kimo: Thanks, but at the moment we do not need any more examples of llTargetOmega issues. We have a fix being tested internally, as mentioned before, and are working to get that deployed as soon as possible. Although various infrastructure issues held up this release, we have another build ready to go as soon as it passes testing. /Sidewinder
April 18th, 2008 at 7:35 AM
Hey some of the fixes listed seem excellent. The folding leg-stairs thing sounds good, I’ve seen this happen a lot, its no biggy, just a question of aesthetics. The phantom floor, I’ve seen happen a lot
@ Prospero, thanks for the sterling service, and taking the time out to reply to comments on the message board as well as upgrading the server patch (you must be busy)! Looking forward to the new kit arriving…..any chance of some nerdy equipment photos?
Also I may be off topic, but could you confirm what the higher build height actually means to us ressys? My build tools still only reach 768m on the Z, will it eventually go up to 4096?
April 18th, 2008 at 7:46 AM
Some people can log in , some can’t. I can accept a TP but I cannot TP myself to any location since yesterday and I have latest client. I keep getting.. try again later… 24hrs later, no change.
I run a business and yes, I have noticed business has been affected in all malls i rent from and at my main location.
April 18th, 2008 at 7:55 AM
llMoveToTarget have the same issue that llTargetOmega, the 2 codes stop to work at the same time.
April 18th, 2008 at 8:02 AM
Re: the higher building limit, I think it will be 4096, yes, but it will also require a client upgrade so that it’s supported on both ends. I’m not sure when that change is slated to go into the client.
April 18th, 2008 at 8:02 AM
Wow. Sidewinder , this looks like you’re not merely talking to us, but actually talking WITH us
WTG in getting back our confidence in u guys. Same goes for having comments open on the ‘trouble’ posts lately. There’s been venting in the comnents but no bash tirade broken pose as might have been expected. We are in this together and by communicating and connecting we wil soon have this rocky road behind us.
April 18th, 2008 at 8:03 AM
OMG!! I just logged on to the sim I spend most time in (Achilles Island), looked to see that we had recieved the update (1.20.1.85162), and….the problem is still there!!! Just as bad, if not worse, than before. Yes, this is an openspace sim.
You can imagine how disappointed I am, after waiting 2 weeks for this deploy, and telling all my friends to just be patient and wait for the patch, the problem had been fixed, to logon and find out it hasn’t been fixed, or even, it seems, mitigated. So amazingly disappointed.
April 18th, 2008 at 8:14 AM
Well, if anything, someone’s doing *something* to al least try to fic stuff :/
Prospero and Sidewinder, it ain’t fun to be you guys, thanks for at least RESPONDING to these posts. Viz. yesterday’s debacle when there wer 150 posts and no replies….. and keep at it, I have seen the othere virtual worlds and SL is unique. Hope it sticks around…
April 18th, 2008 at 8:17 AM
@74 Shin: What problem are you talking about? Is the resolution listed above? I can’t tell what problem to address, because you have said nothing specific. /Sidewinder
April 18th, 2008 at 8:17 AM
@72 Prospero - the increased build limit is already working in the latest RC client 1.20.
April 18th, 2008 at 8:17 AM
Shih — I just checked the simulator node where your island is running. The memory use of all the processes *is* lower, and the machine is not thrashing due to being low on memory as openspace machines have been for the last two weeks. So, it’s not exactly the same problem. However, I do see that several of the regions on that machine do have a lower-than-expected frames per second.
I’ll pass this on to Sidewinder and the rest of the Havok4 team. (It’s possible this is not a Havok4 issue too, but since they’ve already been chasing this it’s a good place to start.)
April 18th, 2008 at 8:19 AM
I’ve noticed a slower rendering on spheres going round
April 18th, 2008 at 8:19 AM
update: it actually happens while trying to put the sime over a parcel-border, i have not found this mentioned on jira.. but i’m really not firm with jira, and i don’t want to mess things up unnecessarily.
Can’t reposition — permission denied
on trying to move a sculpted prim over a parcel-border, before and after the update today, if someone wants to check this.
April 18th, 2008 at 8:25 AM
just tried the new update
-Boats still not float
-it you stand up it kick you still in a corner of the sim
April 18th, 2008 at 8:32 AM
prim over a parcel border, Hmm sure you have object entrie rights in the parcel your trying to move the prim into?
April 18th, 2008 at 8:37 AM
I’m no great ‘techy’ and while using SL FOR 5 months I have been tempted to leave. I understand how complex the demands of the resident population must be, especially when taking into account the vast variations in IT skills and comprehension that must exist. These variables are interfacing the SL software second by second 24 hrs a day. As someone overjoyed when I first ran a dot matrix printer on a 4Mb Ram IBM Pc (way back when!) its hard to believe SL works at all! I share the excitement about the future and SL … so can’t envisage putting it down!
April 18th, 2008 at 8:44 AM
Thank you. A more detailed description has been posted in the JIRA for SVC-2005. The lag behavior looks exactly the same to me viewing the stats window as it has done for the last couple of weeks. If the server is running without the threshing, then maybe that was not all of my problem. Just similar enough to be covering it up. Any of you are quite welcome to drop in and take a look yourselves at what this means to, pardon the term…., gameplay.
April 18th, 2008 at 8:44 AM
@#82 yes, i am sure. the problem persisted the rolling restart and persists a relog with/without having the full rights enabled (right group activated).
April 18th, 2008 at 8:52 AM
@85, ok thats just wierd, I cant seem to reproduce this though Hmm
April 18th, 2008 at 8:55 AM
@74 Shin: I just spent fifteen minutes on Achilles Island, and the simulator was running 90-100% of full speed the whole time, with the exception of one blip to slightly lower. I am not sure which problem you are referring to, and would be happy to take a look at it if we can reproduce the issue. /Sidewinder
April 18th, 2008 at 8:56 AM
Help the fix is out ..now we’re fixing the fix to fix the fix that jack built
April 18th, 2008 at 8:59 AM
@84 Shin: There are various things that can cause a simulator to occasionally “blip” in FPS. The hosts and simulators, as with any complex piece of code, do real work all day long. Sometimes this work is on non-obvious tasks. These performance blips existed before Havok4. The cyclic performance problem noted and resolved with this issue was far more severe and a consistent, obvious drag on movement - as opposed to something that you would notice primarily from watching simulator stats. /Sidewinder
April 18th, 2008 at 9:04 AM
@89 Sidewinder: On Kikai Enkai (was heavily affected before) it looks lots smoother now. While before we had these performance hits every 2 or 3 minutes, and lasting for up to 10 seconds each time, we now have maybe one hit every 10 minutes, lasting only a second or less, not really noticable when not looking at the stastistic bars. Culprit still seems to be physics FPS.
April 18th, 2008 at 9:05 AM
While I do value the responses being keyed into (in all too few of) the blogposts, I can’t help noticing it reduces the possibilities for many customers.
Isn’t there a way to disable the count function for responses from the Moderators and Manipulators to keep the count of 150 open for the customers only?
April 18th, 2008 at 9:09 AM
@91, not really, thats a ‘feature’ of WordPress Mi’thinks
April 18th, 2008 at 9:10 AM
This last month has been a complete nightmare. I understand that Linden needs to keep improvements, latest and greatest and all that jazz. However, many of us have RL SL business going on, pay hefty prices to be here and have poured hours of time, the dollar amount associated with that is “priceless”. Let’s get stable shall we… Develop on the sidelines, test elsewhere until you get it… This reminds me of Microsoft everal years back rushing out the product full of bugs… There is a big difference here. If it aain’t broke don’t fix it… Heard that one before… YES we all want it better, faster, more robust but not in the order of this last month. We, me group are looking at this platform for RL business and at this point that is out of the question. There is no reliabilty to count on and RL business can not work that way. I would be out of business before I begin. This may sound a bit negative and critical however take it for what it is intended… Give us a little peace for a bit… Then let us know with more advance notice so that some of us who plan weeks out know when we are going to subject to the wrath of SL hick ups and plan around them… Thanks
April 18th, 2008 at 9:10 AM
Sidewinder, give me some credit here. I’m not an idiot. I understand that the stats can be read many ways. But…. before the H4 rollout the sim was stable. No lag. I watch this carefully, and several other residents commented on it. After H4 it went into the behavior you just saw yourself if you were watching, and if you had attempted to move during that time you would have seem the effect. The fact that after reading JIRA I assumed that this is what had been causing my problem is an obvious assumption. An openspace sim and cyclical fps problems fit the description in both the blogs I was reading and in the JIRA, and in fact, the ticket opened on the issue by Cliff Commons was closed as a duplicate and rolled into SVC-2005 (it was SVC-2012).
This is not an “occasional blip”. Everything has an “occasional” blip. This is regular and consistent, with never more than 3 or 4 minutes going by without a period of heavy lag. E