Deploys for the week of 2014-06-02 (Updated @ 2014-06-04 17:08)

[Maestro Linden]

UPDATE @ 2014-06-04 17:08: Due to an issue found on all channels, all four channels will be rolled starting now.  See http://status.secondlifegrid.net/2014/06/04/post2285/ for status details.

Release notes for the 2014-06-04 update:

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_Server/14#14.06.04.290645

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_BlueSteel/14#14.06.04.290645

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_LeTigre/14#14.06.04.290645

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_Magnum#14.06.04.290645

 

Second Life Server (main channel)

The main channel will not be rolled this week.  However, please be aware that grid-wide maintenance is scheduled for Tuesday morning (when the main channel would typically be rolled).  This maintenance will cause some temporary disruption to the grid, as detailed here:

http://status.secondlifegrid.net/2014/05/30/post2281/

 

Second Life RC BlueSteel:

BlueSteel will remain on the inventory update project, and will not be rolled this week.  This project enables support for a new version of the inventory service, AISv3.  To make use of this new feature, login with the Sunshine RC Viewer.

 

Second Life RC LeTigre:

LeTigre is moving to a new "group ban" project.  As the name implies, this project adds the ability to ban users from groups.   See the release notes for details about this feature.

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_LeTigre/14#14.05.30.290506

Scheduled Wednesday 2014-06-04 07:00-10:30 PDT

 

Second Life RC Magnum:

Magnum remains on the 'Experience Tools' project, this week.  This week's update includes some minor changes related to experience tools.

https://wiki.secondlife.com/wiki/Release_Notes/Second_Life_RC_Magnum/14#14.05.29.290425

Scheduled Wednesday 2014-06-04 07:00-10:30 PDT

 

We will be monitoring this thread as the code gets released, so feel free to note any observations you have about the server updates.  If you have a specific bug you'd like to report, please file a Jira.

[Darien Caldwell]

Anyone know why this is happening?

http://status.secondlifegrid.net/2014/06/04/post2285/

---

[Posted 3:43 PM PDT, 04 June 2014] We will be performing rolling maintenance for all regions on the main Second Life server channel and all RC channels at approximately 05:00 PM PDT tonight, 04 June 2014.

---

[Ayesha Askham]

Darien

Without any confirmatory comment from a Linden, my best guess is that there is a security exploit that has been found and must be eliminated asap.  Routinely Lindens do not discuss such issues for obvious reasons, but I agree it would be nice to know what they need to pull an emegency all-nighter for.  There has been a spate of serious griefing attacks (ahh shades of 2009), so that may be the reason.

[Perrie Juran]

Are the unscheduled restarts still ongoing?

As of this posting they do not show resolved?

[Theresa Tennyson]

There are starting to be a small number of reports of people no longer able to see baked avatars as of the last couple of days despite their being able to see them properly earler. Has anything been changed in the avatar baking pipeline recently?

http://community.secondlife.com/t5/Second-Life-Server/SSA-messed-up/td-p/2740564

http://community.secondlife.com/t5/Technical/Please-help-Using-LATEST-firestorm-viewer-with-CLEAN-install-i/qaq-p/2741480

[Maestro Linden]

I've posted release notes for yesterday's roll, which finished around 01:30 PDT.

 

Theresa, there haven't been any recent changes to the avatar baking backend.  I would recommend filing a Jira about this baking issue, with attached viewer logs from a session where baking fails.

[Steph Catseye]

If our payments to you would directly linked to the service you provide would you keep deploying  this crap? For how long are we supposed to swallow the fact that every single week for 3 months the same issues that been fixed over a period of 3 days the week before re-appear? Mistakes happen, if they happen every single week we are dealing with unstoppable idiots. How about you don't roll out and fix what has to be fixed once and for all. Compared to REAL and proper webhosting we are paying outragous prices, it is time to provide the service we pay for.

[Redath Pearl]

I see the same happening suddenly again...baking issues of avatars. I thought we had that issues long time behind us....as they were fine for a long time now. A lot of my friends also complain that when they log in - they stay a cloud, and only a avatar test help with it...

[MBeatrix]

There seems to be somewhere a mess-up, but it's a bit too many mess-ups in a short time. This took away from me all the joy of Second Life.

I remember that in the beginning of 2012 there were quite a lot of people ghosted. They were reported as being inworld but were not, hence they couldn't log in. I also remember that the Lab never admitted it was a mess-up, they just told people to keep trying to login repeatedly until they could actually get inworld. The Lab fixed it, somehow, but I never read a single word that could be taken as admission of a mess-up on their part. This whole issue starts to seem like that one, although the effect is different.

[Perrie Juran]

---

MBeatrix wrote:

There seems to be somewhere a mess-up, but it's a bit too many mess-ups in a short time. This took away from me all the joy of Second Life.

I remember that in the beginning of 2012 there were quite a lot of people ghosted. They were reported as being inworld but were not, hence they couldn't log in. I also remember that the Lab never admitted it was a mess-up, they just told people to keep trying to login repeatedly until they could actually get inworld. The Lab fixed it, somehow, but I never read a single word that could be taken as admission of a mess-up on their part. This whole issue starts to seem like that one, although the effect is different.

---

Um, really now? 

Ghosting was the subject of several JIRA's and it took a couple of weeks to root out the cause and apply a fix.

When there was a second flare up further JIRA's were acknowledged and a fix applied.

An example of a public announcement is here:

http://status.secondlifegrid.net/2012/01/28/post1553/

 

So far today I have not had another bake fail so I am unable to start a JIRA myself for that.

 

 

 

[Whirly Fizzle]

---

Theresa, there haven't been any recent changes to the avatar baking backend.  I would recommend filing a Jira about this baking issue, with attached viewer logs from a session where baking fails.

---

Possibly related (?) issue at BUG-6269 where several users report map tiles suddenly failing to load.

The secondlife.log attached to that issue shows some really weird network flakiness including...

2014-05-20T19:45:45Z INFO: LLCurl::completedRaw: Failed to deserialize LLSD. https://sim10335.agni.lindenlab.com:12043/cap/043e06a0-92ba-ee5e-6fe1-0de337f1b8ed [503]: Failed to request agent parameters

2014-05-20T19:45:45Z WARNING: errorWithContent: appearance update request failed, status: 503 reason: Failed to request agent parameters code: 0 error: ""

2014-05-20T19:45:45Z WARNING: onFailure: giving up after too many retries

 

EDIT: Nevermind - he attached a log file that was over 2 weeks old.   :smileyindifferent:

[MBeatrix]

Really, Perrie? I can't see there anything else than the GSP saying there was an issue and that it was being investigated.

At the time, I was about one month SL old, was more or less clueless and felt totally hopeless — I was ghosted for over a week. Still, I created an alt account, but it was ghosted too. I should have left SL right then, instead of believing it would get better and invest time and money in it.

As I said, never read a single word that could be  taken as admission of a mess-up, never an apology.

About this baking issue, it seems obvious to me that if the Lab doesn't know about it is because they don't know what they are doing. And that's exactly what I feel most of the time: the Lab knows where they want to take SL, but just don't know exactly how to get there, so they keep experimenting.

You bet I'm upset and discouraged.

I'll file a JIRA when I have the time and patience for that. But know that it isn't right, either... I mean, I pay for all this and still I have to go doing part of the Lab's job, trying things to find out what is wrong. It is like going to the clinic that is killing you with wrong prescriptions and tell them what your sympthoms are so they keep experimenting until they kill you for good.

There have been serious issues with the assets service (messed up databases or whatever) since Christmas. Has anything consistent been done about it? Not that I can tell. The issues persist and even get worse once in a while.

I'm sick 'n tired, really.

[MBeatrix]

JIRA filed: https://jira.secondlife.com/browse/BUG-6275

[Theresa Tennyson]

---

MBeatrix wrote:

 

I'll file a JIRA when I have the time and patience for that. But know that it isn't right, either... I mean, I pay for all this and still I have to go doing part of the Lab's job, trying things to find out what is wrong. It is like going to the clinic that is killing you with wrong prescriptions and tell them what your sympthoms are so they keep experimenting until they kill you for good.

---

Actually that's a perfect metaphor. The clinic isn't trying to kill anyone, and the prescriptions generally work for most people. But some people have complications that the clinic needs to know about and if those people just go home and sulk instead of telling the clinic what the complications are, they will die.

There is little that can happen in Second Life that can affect only a specific part of the world. There isn't a switchboard at the Second Life servers with a slot labled "Portugal." All data is treated basically the same way and then it's picked up by people's ISP's through their pre-made arrangements with a variety of "backbone" services.

 

When server-side appearance first rolled out it turned out a small number of people were using services that handled the data in an unusual way and it didn't work for them. After people started JIRA entries and sent in data, eventually a fix was found that worked for them. It sounds like your ISP has changed the way it handles some data and that causes a problem that wasn't there before. The problem is probably fixable but not if Linden Lab doesn't know what the problem is. And Linden Lab is not in Portugal and doesn't use your ISP - doesn't even know who your ISP is - so they can't see this problem directly. So you need to give them information about your specific situation in this case or its impossible for them to do anything.

 

 

[MBeatrix]

It sounds like your ISP has changed the way it handles some data and that causes a problem that wasn't there before

Sorry, Theresa, but you're just guessing like everyone else — me included.

However, this is typical: the ISP says they didn't change anything, and SL Support told me to talk with my ISP. But SL is much more subjected to changes than any ISP. Also, I'm not sure it is a network problem, since all sculpties, mesh and its textures bake fine — most probably isn't, but I'm guessing again, which is all I can do.

Yes, it is true that we must provide the sympthoms to get a cure, eventually. But I seriously doubt that it is possible to survive when the disease was induced by the doctors in the first place and they just keep experimenting things on you, not knowing exactly what they are doing.

Anyway, the JIRA was filed.

 

P.S. - But let me tell you a bit more how I feel about the JIRA... More than one year ago I filed a JIRA reporting a Pathfinding bug that was marked as duplicate and closed. Months later, the Lab announced it had been fixed, but it wasn't totally, or what they fixed was a different bug. Anyway, the bug I reported is still active, "eating" some regions' memory and causing the need to have them restarted almost on a daily basis, depending on what happens there. About one month ago, Support finally told me they are working on it.

Once in a while I think that some people spend more time in the forums than they do inworld, actually trying to do anything there, since Second Life is always so flawless in their own words (no, I don't think this applies to you.)

[Perrie Juran]

---

MBeatrix wrote:

Really, Perrie? I can't see there anything else than the GSP saying there was an issue and that it was being investigated.

At the time, I was about one month SL old, was more or less clueless and felt totally hopeless — I was ghosted for over a week. Still, I created an alt account, but it was ghosted too. I should have left SL right then, instead of believing it would get better and invest time and money in it.

As I said, never read a single word that could be  taken as admission of a mess-up, never an apology.

About this baking issue, it seems obvious to me that if the Lab doesn't know about it is because they don't know what they are doing. And that's exactly what I feel most of the time: the Lab knows where they want to take SL, but just don't know exactly how to get there, so they keep experimenting.

You bet I'm upset and discouraged.

I'll file a JIRA when I have the time and patience for that. But know that it isn't right, either... I mean, I pay for all this and still I have to go doing part of the Lab's job, trying things to find out what is wrong. It is like going to the clinic that is killing you with wrong prescriptions and tell them what your sympthoms are so they keep experimenting until they kill you for good.

There have been serious issues with the assets service (messed up databases or whatever) since Christmas. Has anything consistent been done about it? Not that I can tell. The issues persist and even get worse once in a while.

I'm sick 'n tired, really.

---

I'm really not going to argue with you here.

If you are wanting them to say, "We're sorry that SL hasn't lived up to your expectations," or that every time someone has a problem they say, "Oh, we are sorry, we screwed up," that probably isn't going to happen.

I've had my problems also, one of them was major.  There was a tiny, tiny number of us who had the problem.  And it got fixed.  

I looked at your Viewer log.  I really don't know enough but I'm pretty sure the question is going to be, 'Why are you having so much trouble connecting to the Server?"  It's possible the failure is with something LL did.  Or the failure could be elsewhere.

 

 

[MBeatrix]

I'm really not going to argue with you here.

I'm not looking for arguments, either, Perrie, just letting people know — especially the Lab — how I feel about it.

I've had my problems also, one of them was major.  There was a tiny, tiny number of us who had the problem.  And it got fixed.

 There seems to be a long history of "per user" issues, problems that affect only some people. Why does that happen? One of the Lab's mysteries...

I looked at your Viewer log.  I really don't know enough but I'm pretty sure the question is going to be, 'Why are you having so much trouble connecting to the Server?"  It's possible the failure is with something LL did.  Or the failure could be elsewhere.

Yes, I saw that too, and I probably know much less than you do... So I'm not excluding anything. But, again, "per user" issues do happen when the right condtions of each one of them trigger whatever. Nevertheless, the present issue seems to affect some people in mainland Europe only, especially Portugal, as far as I know.

 

One more thing, that may be related or not: since Christmas I've been experiencing inventory issues, and each time I unpack anything Firestorm starts fetching inventory and won't stop until I re-log.
I'm sure you remember all those inventory database unscheduled maintenances from Christmas till a bit after New Year's celebrations. May mean something or nothing at all... But one thing is certain: the assets service isn't working well.

[Malate Banufong]

Currently, on my world map, a lot of sims are showing offline (see Blake Sea area). Is this a problem with the world map, or are the sims being restarted because of a problem. There isn't anything showing on the girid status page.

 

OK, yes it is the world map showing offline sims when they are online.

[Perrie Juran]

---

MBeatrix wrote:

I'm really not going to argue with you here.

I'm not looking for arguments, either, Perrie, just letting people know — especially the Lab — how I feel about it.

I've had my problems also, one of them was major.  There was a tiny, tiny number of us who had the problem.  And it got fixed.

 

There seems to be a long history of "per user" issues, problems that affect only some people. Why does that happen? One of the Lab's mysteries...

I looked at your Viewer log.  I really don't know enough but I'm pretty sure the question is going to be, 'Why are you having so much trouble connecting to the Server?"  It's possible the failure is with something LL did.  Or the failure could be elsewhere.

Yes, I saw that too, and I probably know much less then you do... So I'm not excluding anything. But, again, "per user" issues do happen when the right condtions of each one of them trigger whatever. Nevertheless, the present issue seems to affect some people in mainland Europe only, especially Portugal.

 

One more thing, that may be related or not: since Christmas I've been experiencing inventory issues, and each time I unpack anything Firestorm starts fetching inventory non stop.

I'm sure you remember all those inventory database unscheduled maintenances from Christmas till a bit after New Year's celebrations. May mean something or nothing at all... But one thing is certain: the assets service isn't working well.

---

As far as "per user" problems go I can tell you that SL is far from alone in this.  At one time I got so tired of hearing people say "I never have a problem with this game or that game," that I went and checked their message boards.  And they were ripe with "per user" issues.

SL maybe has a bit more but I'm not really sure.  What I do know is that SL has to do things those other services do not because SL is User Created Content and the number of real time updates far exceeds anything those other services need to handle.  SL has to send you a ton of data that other services can simply preload on your computer.

I'm also aware of the Asset Server failures.  Hardware can fail and do it unexpectedly.   LL has acknowledged there are problems with the asset systems.  LL has chosen to address that step by step.  Maybe it is and maybe it isn't the the right way to handle it.  Maybe what they need to do is shut down for a few months and retool.  But in the meanwhile they are doing it in real time and that must make it much more tedious to do.

[MBeatrix]

Yep, you got quite a few points there, Perrie.

LL has chosen to address that step by step.  Maybe it is and maybe it isn't the the right way to handle it.  Maybe what they need to do is shut down for a few months and retool.

The thing is, Second Life is a business, and I don't think the Lab can afford shutting it down for a few months.

Anyway, there are some developments in the JIRA. It seems the bake service isn't allowing us connecting to some parts of it. Per country? Per ISP? We'll see...

[Perrie Juran]

---

MBeatrix wrote:

Yep, you got quite a few points there, Perrie.

LL has chosen to address that step by step.  Maybe it is and maybe it isn't the the right way to handle it.  Maybe what they need to do is shut down for a few months and retool.

The thing is, Second Life is a business, and I don't think the Lab can afford shutting it down for a few months.

Anyway, there are some developments in the JIRA. It seems the bake service isn't allowing us connecting to some parts of it. Per country? Per ISP? We'll see...

---

Glad to see you are getting action on your JIRA!

I'll be keeping my fingers crossed for you.

[MBeatrix]

Thanks, Perrie. :matte-motes-bashful-cute-2:

I had to make the JIRA non public, since some of the tests may show my IP adress and I don't want to give some griefer the chance of trying anything. But I'll keep you posted here how it is going. For now, I'm just doing tests that Monty Linden is suggesting, and I got no conclusion from them so far, except it is not a viewer issue, which I already knew, since I tried 3 different viewers.

[MBeatrix]

And... Can be an ISP problem (or ahead) it seems — bad DNS. Which is easily fixed using Google public DNS servers.

It is easy to rant against the Lab, but Monty Linden was just great, so I owe an apology to them all this time. Noted. And a THANK YOU SO MUCH!

Someone, somewhere changed something. So I have someone else to rant against as soon as I find out who. :matte-motes-wink-tongue:

[Perrie Juran]

---

MBeatrix wrote:

And... Can be an ISP problem (or ahead) it seems — bad DNS. Which is easily fixed using Google public DNS servers.

It is easy to rant against the Lab, but Monty Linden was just great, so I owe an apology to them all this time. Noted. 

And a THANK YOU SO MUCH!

Someone, somewhere changed something. So I have someone else to rant against as soon as I find out who. :matte-motes-wink-tongue:

---

Very glad to hear you got it fixed.

I've been involved three times working with a Linden on a problem and each of them really was wonderful to work with.

 

[MBeatrix]

Thanks, Perrie, I'm very glad too. :matte-motes-big-grin-squint:

Quoting Monty Linden: "You have bad DNS information for our bake service. Instead of talking to Linden servers, the viewer and your browser have been trying to connect back to your own computer. If 'xxxxxx.pt' is your ISP, the problem may be there. Show them the nslookup command and output and they should understand the problem."

And again: "There are some odd cases that arise when talking to the baking service which is why I hopped on this. They haven't all been resolved. But ISP problems are still far more common and, honestly, probably becoming more common."