Update 2008-06-07 13:51 : the rolling restart is done. Less than 100 regions remain down; we will get those up as soon as possible. Uf-da!
Update 2008-06-07 13:16: We took another little break as the database decided it didn’t really want to talk to us for a few minutes. Once its tantrum was done, we’ve started up again. The rolling restart is about 95% complete, and when this one’s done, boy howdy are we going to be relieved!
Update 2008-06-07 12:08 : The rolling restart is about 90% complete.
Update 2008-06-07 10:46 : The rolling restart is about 65% complete.
Update 2008-06-07 09:52 : The rolling restart is about 50% complete.
Update 2008-06-07 08:53 : The rolling restart is about 37% complete. We’ve paused it because the central database has briefly become overloaded. It will begin again shortly. Once again, we apologize for the bug in the previous version that is causing many regions to stay down for a couple of hours during this rolling restart; we are doing what we can to get them back up.
Update 2008-06-07 07:52 : The rolling restart is about 25% complete. A fair number (several hundred) of regions are down, and have been down for >1h. We’re working on getting them back up right now.
Update 2008-06-07 06:33 : The rolling restart is about 9% complete. As with yesterday, because of a bug in 1.22.1 that this release fixes, some regions are taking a very long time to shut down, and will be down for more than the usual we’ve seen in rolling restarts. Some regions will be down for more than an hour.
Update 2008-06-07 04:57 : The rolling restart to deploy 1.22.2 to the second half of the grid will begin shortly. NOTE: we are aware of and working on SVC-2495. We recognize that this is causing pain for people who’ve received the update already, and will cause pain for people who have received the update. However, we need to do this update for stability reasons. Will will issue more information about SVC-2495 as it is available.
Update 2008-06-06 07:30 : Network problems continue. We are postponing the second half of the rolling restart to tomorrow (Saturday) morning at 5AM.
Update 2008-06-06 06:21 : The second-half rolling restart has not yet begun, as we are still being plagued by some network problems.
Update 2008-06-05 06:18 PM : If land is set to group build only, you must be a member of the group *with the group activated* to rez objects on the land. If you do not have the group activated, you will get a misleading error message about not being allowed on the parcel.
We are investigating and will address in a future version.
Update 2008-06-05 12:54 : The first-half rolling restart is complete. A handful of regions remain down, and we are now working on getting them running.
Update 2008-06-05 10:54 : The first-half rolling restart is 75% complete.
Update 2008-06-05 09:49 : The rolling restart has started up again. Because of the interruption, unfortunately a number of regions that were already restarted once are going to be restarted again.
Update 2008-06-05 09:20 : The central database machine has become overloaded, and that affects rolling restarts along with everything else. The rolling restart has been paused; there are some regions down that will stay down until the database recovers.
Update 2008-06-05 09:00 : The first half rolling restart is 57% complete. However, the memory leak bug that’s fixed in this rolling restart of course still affects the simulators we’re shutting down, and as a result it’s wreaking some havoc (not havok) with the RR. Quite a number of regions are stuck down for 1-2 hours as a result of this. We apologize for this inconvenience, and are working on it.
Update 2008-06-05 07:57 : The first-half rolling restart is 28% complete.
Update 2008-06-05 06:59 : The first-half rolling restart is 8% complete.
Update 2008-06-04 06:35 : Because of the memory leak issue that is fixed in this deploy– but exists with what’s running on the grid right now– some regions are taking a very long time to shut down, and may be down longer than usual during a rolling restart.
Update 2008-06-05 06:07 : The rolling restart of the first half of the grid began at 06:00.
Update 2008-06-04 20:30 : The deploy of 1.22.2 to 186 regions is now complete.
We have deployed version 1.22.2 of the server code to the “Second Life Beta Server: channel on Aditi, the Preview Grid
Fixes include the following:
- SVC-1937: Avatar no longer gets stuck in hover when walking up stairs or impacting prims at certain angles
- SVC-2014: Avatars flying within the bounding box of a physical object no longer slows the physical object
- SVC-2443: Flying from hover is no longer “sticky”.
- DEV-14174: Avatar no longer sinks slowly while hovering (and ‘hopefully’ no longer displays other slow rise/fall symptoms in hover modes)
- DEV-15875: Avatar now collides with tall prims, rather than walking up walls
- Internal: Resolved several crash modes
- Internal: Resolved server memory leak
Over the next three days, this code will be deployed to Second Life in a rolling restart. Our plan is as follows:
- Wednesday, 06/04 morning: Proceed with QA validation of the above fixes/changes today on Preview Grid
- Wednesday, 06/04, 7:30PM: we will do a pilot rolling restart to about 150 regions
- Thursday, 06/05, 5:00-9:00AM: we will do a rolling restart of half of the grid
- Friday, 06/06, Saturday, 06/07 , 5:00-9:00AM: we will do a rolling restart of the rest of the grid
As with all rolling restarts, once the restart is announced for a given region, there’s no way to delay it. Each region will be restarted only once for the deploy, and should be down for 5-10 minutes. If your region is down for more than 30 minutes, please contact support.
To log into the Preview Grid, you can use the same Second Life viewer you already use to log into Second Life! At the login screen, hit CTRL-SHIFT-G. At the bottom of the screen to the right of the “Quit” button, you will see a dropdown widget that allows you to select the grid that you want to log in to
Select “Aditi” to log into the Preview Grid. (A number of other grids are listed in this dropdown. These are internal development grids which are not available for public access.)
Please report any new issues in the Issue Tracker and be sure to set “Affects Version/s” to “1.22.2” (which is at the bottom under “Unreleased Versions”). When reporting a problem, please give as much information as possible: what region you were in, exactly what you did and what behavior you saw, when it happened, and the version of both your viewer software and of the server software running in the region where you saw the problem.