Sign in to follow this  
Oskar Linden

Deploys for the week of 2012-09-24

Recommended Posts

There is no promotion to main channel this week. Since there was no RC rollout last week we do not have anything to promote Tuesday morning. There will be no rolling restart.

UPDATE: There were blocking bugs found in both the RC's planned for release this week. There will be no releases Wednesday morning. There will be no rolling restarts.

Second Life Server (main channel)

This code will stay the same. Regions will not be restarted.

 2012-09-25, 5:00am: Release Notes: https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_Server/12

  

Second Life RC BlueSteel

This code will stay the same. Regions will not be restarted.

2012-09-26, 7-11:00am: Release Notes: https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_BlueSteel/12

 

Second Life RC LeTigre

This code will stay the same. Regions will not be restarted.

2012-09-26, 7-11:00am: Release Notes: https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_LeTigre/12

 

Second Life RC Magnum

This code will stay the same. Regions will not be restarted.

2012-09-26, 7-11:00am: Release Notes: https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_Magnum/12

 

We will be monitoring this thread during the next week so please feel free to post issues that you feel have been introduced by the new code. Please file a JIRA for issues you find and post the JIRA link into this thread. It really helps us out. When determining if issues are relevant or not research is key. Tracking down exactly the right situation where an issue is occurring greatly speeds up the development process to get fixes in place.

I appreciate your help. Have a good week!

 

__Oskar

 

p.s. If you are interested in helping test SecondLife in beta please join the group "Second Life Beta" in-world. We also have an email list where we communicate upcoming projects and how you can help. (https://lists.secondlife.com/cgi-bin/mailman/listinfo/server-beta ) Once a week we meet on ADITI to discuss new features, new bugs, new fixes, and other fun stuff. You are more than welcome. Information is here:https://wiki.secondlife.com/wiki/Server_Beta_User_Group

 

This is not a forum thread for off-topic and unproductive ranting. I will no longer tolerate it. This forum thread is to discuss newly found issues in the Release Candidate channels. If you feel the need to have your opinions heard on topics unrelated to new bugs then please start your own thread. If an individual continues to post unproductive and off-topic comments I will take measures to remedy that.

If you are unsure about your posting please read this : "Linden Lab Official:Community Participation Guidelines"

Share this post


Link to post
Share on other sites

potholes_003.png

 

OK, I know this may be an oddity, but potholes on the Mainland Roads?

I ran into this walking down a road at a SIM crossing. 

Sometimes you can 'walk out' of it, other times takes a relog.

It only occurs walking South.

Or is this a problem for the LDPW?

You are at 286,253.2, 265,471.4, 92.6 in Huineng located at sim7519.agni.lindenlab.com (216.82.35.12:13000)
Second Life Server 12.09.07.264510

 

 

Share this post


Link to post
Share on other sites

I poked around this general area and noticed a build (and I use the term very loosely) in the sky that actually straddles the road in the southeast corner of Cipactli, which is right around this problem area. You can see it on the map too. You should contact LDPW - this may be causing the problems. I had some problems with border crossings in the area too but they seemed temporary.

Share this post


Link to post
Share on other sites

We found blocking issues during our pre-RC smoke tests. These issues will block the Wednesday morning RC releases. Regions will not be restarted.

__Oskar

Share this post


Link to post
Share on other sites


Perrie Juran wrote:

potholes_003.png

 

OK, I know this may be an oddity, but potholes on the Mainland Roads?

I ran into this walking down a road at a SIM crossing. 

Sometimes you can 'walk out' of it, other times takes a relog.

It only occurs walking South.

Or is this a problem for the LDPW?

You are at 286,253.2, 265,471.4, 92.6 in Huineng located at sim7519.agni.lindenlab.com (216.82.35.12:13000)

Second Life Server 12.09.07.264510

 

 

This looks like one that LDPW can fix. You will notice that at many sim crossings, there are transparent prims just under the roadbed, extending across the boundary. These are there to make up for an ancient problem, prims can only be solid in one region at a time. I don't see any in this spot. Give them the coordinates in a ticket, tell them about your "pothole" experience, and mention that this could be a good spot for buffer prims.

Share this post


Link to post
Share on other sites

I don't know if this is the correct place to post about this new bug, reproduced by Whirly, and which I believe must have arisen in last week's rollout. I have filed a bug report.

When my avi, who is seated in a chair animated by a sit animation, attaches and/or detaches an object with an eating animation in it, she pops up in a sitting position like this, and the sit animation stops, tho sometimes if the popup happens on attach instead of detach, the eating animation will continue. Also notice the arm not eating will sometimes do odd things:

 

Screen shot 2012-09-26 at 12.16.51 PM.png Screen shot 2012-09-26 at 12.25.19 PM.png
This is on Firestorm, Phoenix, and Viewer 3, and on all four of my sims. It happens with chairs with all three of the sit scripts I have used over a four year period. It is recent, since I discovered it myself while making new content and not from customers -- I have sold tens of thousands of items with this configuration of animations. (And I currently have two sims filled with them.)
 
One thing I notice is that I do not get a permission to animate request, as I do when the object is rezzed rather than worn. That may be normal.
 
 

Share this post


Link to post
Share on other sites

Pamela

This is something that I have also seen, but it is not new.  Also there is a similar though somewhat different bug that occurs if the av is being animated by a sit and the av types something in local chat.  Again the av pops up, sometimes into a default sit, sometimes continuing the animation, but risen above the location of the poseball by about 0.5m

Maestro told me that since it was intermittent (which it has been until recently), they could not repro it and the JIRA I raised was closed.  That was some 12 months ago now.

Whether a new BUG report should be filed I do not know, but I do know that this issue has been around in one form or another for a very long time.

Share this post


Link to post
Share on other sites

I've seen this happen when first creating some of my first animations, when triggering them in reverse order than they are normally used in, It normally can be caused by part of the animation not being adjusted when created, that is even if you want an animation to do nothing you still have to adjust every thing at least a bit or it might get over ridden by an other animation, so if an arm was the only thing not adjusted then it would be the only part to get moved by an other animation.  All so animation priority could be an issue specially if the animation are triggered in reverse order.  Sounds like you didn't have a problem before last week, but the things above may help point to where the bug is.

Share this post


Link to post
Share on other sites

I have been using this same configuration of animations for almost five years. I have 68 dining sets listed on the marketplace, almost a whole sim devoted to them. I have never seen this bug until yesterday, when it was easily reproduced by Whirly Fizzle, who first responded to the bug report -- 7/10 times she tried.  That constitutes major breakage. Anything more than 1/10 does, IMO.

The animations are two extremely common ones, by Craig Altman and Robin Sojourner. 

 

 

Share this post


Link to post
Share on other sites

Maestro has reproduced the effect or some variation of it, with other animations, which fail to behave as expected 50% of the time for him. The bug priority is Major.

Share this post


Link to post
Share on other sites

Several Mainland sims I frequent are in very deep lag right now, with a couple of different patterns--with or without elevated Sleep Times:

Anilis, Time Dilation 0.4 - 0.5, Script time over 40ms; Time Details show Sleep Time > 20ms, but 0 Spare Time.

Tuliptree: Time Dilation about 0.5, script times over 40ms, Sleep Time > 20ms, but 0 Spare Time.

Tussock: Time Dilation 0.5 - 0.6, script times > 40ms (sleep time negligible)

Miramare: Time Dilation 0.3 - 0.4, script times > 60ms (sleep time negligible)

EDIT TO ADD another in the long-sleep category:

Eagan: Time Dilation about 0.4, script times > 40ms, Sleep Time > 20ms, but 0 Spare Time.

AND another not very sleepy:

Morris: Time Dilation about 0.4, script times about 50ms, Sleep Time about 1ms but 0 Spare Time.

EDIT AGAIN: Eagan is back to normal now.

Share this post


Link to post
Share on other sites

Really this is more an additional comment but I didn't want to make it an ETA.  It follows on from Pamela Galli's posts.

This issue with broken poses that I mentioned may be more understandable of one remembers the "head-nodding" animation played when one's avatar enters chat.  This being a default animation I would expect it to have a very low priority, being overridden by most poses and AOs.  It is a minimal movement but it may be sufficient when it runs to break other poses.

Perhaps something has upset the animation-priority system, especially with respect to default animations?

That would explain the behaviours that I have seen on too many occasions for it to be something I do wrong.

Share this post


Link to post
Share on other sites


Ayesha Askham wrote:

Really this is more an additional comment but I didn't want to make it an ETA.  It follows on from Pamela Galli's posts.

This issue with broken poses that I mentioned may be more understandable of one remembers the "head-nodding" animation played when one's avatar enters chat.  This being a default animation I would expect it to have a very low priority, being overridden by most poses and AOs.  It is a minimal movement but it may be sufficient when it runs to break other poses.

Perhaps something has upset the animation-priority system, especially with respect to default animations?

That would explain the behaviours that I have seen on too many occasions for it to be something I do wrong.

Maestro says it has to do with the sim not keeping track of how many animations are playing.  He says he does not see the effect with the default animations such as in the Library chair, however.

 

 

Share this post


Link to post
Share on other sites

There remain other sick sims with similar symptoms of lag. I don't know whether it helps to report them here or if I should just file a support ticket as I come across them... or maybe I should just ignore them until somebody else notices and files a ticket, since they don't directly affect me all that much.

I'll list a few here, I guess, just in case it's somehow helpful in diagnosing whatever has caused these regions to go into dilation.

  • Bay City - Rollers : dilation 0.4, sleep time > 20ms
  • Thyris : dilation 0.4
  • Didugua : dilation 0.6
  • Achemon : dilation 0.4, sleep time > 30ms
  • Idia : dilation 0.4, sleep time 30ms
  • Nolidae : dilation 0.6
  • Aglia : dilation 0.4

(Note: Most of these are along the SLRR Main Line because that's mostly where I've looked, presumably not because anything on the rails causes the problem.)

 

Share this post


Link to post
Share on other sites

One does kinda wonder what's causing this. Those frame-length and longer sleep times may arouse suspicion of region idling gone haywire, but only about half of the sick sims have that particular symptom. (And otherwise, that idling stuff has seemed rock solid.)

I don't really think the sims are just pining to be restarted, but if we really are seeing the hairy edge of extended uptime, we could be in for a long bout of this, given the lack of RCs in the pipeline for promotion.

Share this post


Link to post
Share on other sites

I'm standing in Achemon right now which is showing these symptoms. The problem seems to be a rogue script or many rogue scripts that are causing sleep time. Sleep time peaks at just under 28 ms, script time is often over 50 ms, but the scripts run percentage is less than 0.5 percent.

ETA - exactly the same high script and sleep times, low script running percentage in Idia and Bay City - Rollers. All of these sims are on RC servers.

Nolidae and Didugua are showing time dialation, high script time and less than 1% of scripts run but are NOT showing elevated sleep times. They're both on main-channel servers.

Share this post


Link to post
Share on other sites

AFAIK, the main difference between the RC servers and the main channel is that one has been running the same software version longer, so has generally been without a restart longer. Unfortunately, I can't remember for sure which is which now.  I vaguely recollect that the RCs had an attempted update that got rolled back, while there was nothing to promote to the main channel, so the RCs would have been restarted more recently -- but I may be remembering that all wrong.

Share this post


Link to post
Share on other sites

Revisiting this problem, I met an old bug-stomping friend (Battery Street Irregulars lives!).  Of the sims I reported yesterday, all but Idia are back to normal, as is Okeanos that Mari reported. None of the ones from the day before got sick again, and no new ones to report on the SLRR lines, nor Bay City transit-served sims, nor the Dore-Luna trolley path, which is all that my scripts sample. 

My friend found another one at Laserlight.  That's a sim I don't sample with my scripts, so it may well have been this way for a while. It's also a BlueSteel RC sim (of course all these are on 12.09.07.264510), and shows the inflated Sleep times.

One might expect that there'd be new ones cropping up, if it were a simple, gradual memory leak. I can't say that's not happening at all, since I only sample a hundred or so sims, but it does suggest that something else may be the cause.

EDIT: New data point of some significance: Saturnia has the high sleep time, but is the main server release. So the apparent relationship between RC and inflated sleep time seems to have been coincidental.

Share this post


Link to post
Share on other sites

Adding some more from a brief tour through a bit of Route 3 before RL called. Note that these are not known to be new occurrences of the problem, just new discoveries:

  • Serpentata : Main release, high sleep time
  • Nolidae : Main server, 0 Sleep time
  • Dagger : Main server, 0 Sleep time
  • Highflyer : Main release, 0 Sleep time
  • Muslin : LeTigre RC, high Sleep time

Share this post


Link to post
Share on other sites

Qie

Remember, Main server is now running (officially) the same code as the RC channels, since the last promotion to Main from the RCs two weeks ago.

So any issues seen on the RCs may well be seen on Main Server sims, they just haven't been running that code for as long as the RCs.

As to the tie-in between server uptime between restarts and lag-related issues...ie memory leaks: they are bound to have been exacerbated by the long time since a restart for many sims.  The sims that copped the faulty code were Main Server not RCs.  None of the failed RC candidate code got anywhere near an Agni sim according to Oskar.

Share this post


Link to post
Share on other sites

Thanks Ayesha.  Yes, I know the main release and RC channel sims are currently on the same code. And now that I know that the weird sleep effect obtains on some main release regions, too, I've abandoned that whole idea. (I'd been toying with two possibilities: the longer uptime for one or the other--thanks for pointing out that it was the main release sims that had been restarted most recently--and the chance that the RCs might have extra and/or different monitoring code, outside the sim release itself. Anyway, academic now.)

For now, I think I'll drop the subject, unless one of the sims my scripts regularly visit gets sick again. It's an interesting puzzle, but I sorta suspect there's a Linden somewhere who has the solution, so solving it again is kind of pointless.

I have to say, it's a bit disturbing to see so many Mainland sims in such dire condition, with only a few of us noticing. It feels a bit lonely in here.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
Sign in to follow this