Leaves in the dropping of the client GUN8 (now 9) uuid binding message, since this was the much more common case from the viewer-side and this can only affect viewers.
Addresses http://opensimulator.org/mantis/view.php?id=6768
Thanks to talun for the patch on that commit - in the end I took a different approach that also deals with invalid PRIM_LINK_TARGET
However, not yet generating the same warning on invalid PRIM_LINK_TARGET as seen on LL grid
This commit also adds regression tests for some cases of llGetLinkPrimitiveParams()
By not binding UUID -> "Unknown User", we leave open the possibility that the binding may be correctly resolved at a later time, which can still happen in some Hypergrid situations.
Observed behaviour of LL viewer 3.3.4 is that a dropped bind request is not retried until the next session.
Add proxy cache clearing when some properties are changed. This fixes
a problem where objects would stop colliding of they were moved
with setPosition mulitple times.
Instead, "show client stats" now prints "Off!" so that exception is not thrown and we know which entries in ClientManager are in this state.
There's a race condition which could trigger this, but the window is extremely short and exceptions would not be thrown consistently (which is the behaviour observed).
It should otherwise be impossible for this condition to occur, so there may be a weakness in client manager IClientAPI removal.
This is necessary because the hypergrid groups checks (as referenced by estates) require an agent circuit to be present to construct the hypergrid ID.
However, this is not around until Scene.NewUserConnection(), as called by CreateAgent() in EntityTransferModule.
Therefore, if we're dealing with a hypergrid user, delay the check until NewUserConnection()/CreateAgent()
The entity transfer impact should be minimal since CreateAgent() is the next significant call after NewUserConnection()
However, to preserve the accuracy of query access we will only relax the check for HG users.
This is to avoid excessive and inconsistent delays between behaviours that currently need to embed sleeps in other actions (e.g. physics) and other behaviours.
Might need a more sophisticated approach in the long term.
This is not guaranteed to be accurate since users may be left "online" in certain situations.
For example, if a simulator crashes and they never login/logout again.
To counter this somewhat, only users continuously online for less than 5 days are shown.
This occurs on v2 teleport since the source region now waits 15 secs before closing the old child agent, which could still receive chat.
This commit introduces a ScenePresenceState.PreClose which is set before the wait, so that ChatModule can check for ScenePresenceState.Running.
This was theoretically also an issue on v1 teleport but since the pause before close was only 2 secs there, it was not noticed.
This requires a new GridService.GetDefaultHypergridRegions() so ROBUST services require updating but not simulators.
This method still returns regions flagged with just DefaultRegion after any DefaultHGRegions, so if no DefaultHGRegions are specified
then existing configured defaults will still work.
Immediate use is for conference where we need to be able to specify different defaults
However, this is also generally useful to send experienced HG users to one default location and local users whose specified region fails (e.g. no "home" or "last") to another.
If the port is specified it is added but a ":0" is not added if the port is zero.
This enables the hypergrid address short form "hypergridGateway:regionName"
which is handled by the parser but failed because of this zero port addition.
This is very similar to my earlier revert in bcb8605f84 and fails for the same reasons.
Reverting this change because it causes a problem if access is denied to the user.
This reverts commit c7a8afbb8d.
Fallback doesn't work at this level as the change of destination isn't communicated to the source region/viewer
Reverting because this introduces a bug when access does fail.
More detail in revert of main commit.
This reverts commit ec32c1d4b6.
This exception was very likely harmless since it occurred after the restart had taken place, but still misleading.
Thanks to SCGreyWolf for the code change suggestion in http://opensimulator.org/mantis/view.php?id=6747, though I did this in a slightly different way.
Also separates out log messages to distinguish between close not finding an agent and wrong auth token, and downgrades former to debug and latter to warn
This approach does not work - it is taking place too far down the login process where really the region checking
could only be done when the hg map tiles are linked on the main map (messy and probably impossible) or possibly
when the final destination is fetched at the very first stage of teleport (which couldn't be done without a protocol
change to pass the agentID as well as the requested regionID)
This reverts commit 3d9b73c47a.
These would be specified in the [GridService] section of Robust.HG.ini, which already lists these in the example text.
Untested patch so that Neb can easily pull in for testing, though shouldn't disrupt existing hg logins since fallback processing is a bit of code stuck on the end of the login sequence.
If this is attempted, they get a "Try moving closer. Can't sit on object because it is not in the same region as you." message instead, which is the same as current ll grid.
Sitting on ground is okay, since viewer navigates avatar to required region first before sitting.
Currently, first name and last name are input separate but are concatenated with a space in the middle to form a regex.
So to sit all bots with the first name "ima", for instance, the command is "sit user name --regex ima .*"
"sit user name" will currently only sit the given avatar on prims which have a sit target set and are not already sat upon.
Chiefly for debug purposes.
Now, bots will only connect at startup if this switch is specified.
If it is not specified, then a separate "connect" command is required on the pCampbot command line
"quit" console command now requires bots to be separate disconnected first before quitting.
This is to prepare the way for disconnecting/reconnecting different numbers of bots in a pCampbot session.
And hopefully resolves bug where console appears not to be reset if Environment.Exit(0) is called on a different thread
This is because they were wrongly looking for both regionid and region_id parameters in the same request.
Now only region_id is required (and recognized), regionid having been already deprecated for some time.
This is essentially Michelle Argus' patch from http://opensimulator.org/mantis/view.php?id=6737 but with tabs replaced with spaces.
Thanks!
Instead adjusts code with that from Packet.BuildHeader(byte[], ref int, ref int):Header in libomv
This stops packet decoding failures with agent UUIDs that contain 00 in their earlier parts (e.g. b0b0b0b0-0000-0000-0000-000000000211)
Thanks to lkalif for pointing this out.
This message is seen on V2 if one attempts to quickly re-teleport from a source region where one had previously teleported to a non-neighbour and back within 15 secs.
The solution here is for the user to wait a short while.
This message can also be seen on any teleport protocol if one recieves multiple teleport attempts simultaneously. Probably still useful here to help identify misbehaving scripts.
This approach has problems if a client quits without sending a proper logout but then reconnects before the connection is closed due to inactivity.
In this case, the DoNotCloseAfterTeleport was wrongly set.
The simplest approach is to close child agents on teleport as quickly as possible so that races are very unlikely to occur
Hence, this code now closes child agents as the first action after a sucessful teleport.
Normally, avatar motion is zeroed if colliding with a stationary object so
they don't slide down hills and such. Without volume detect check this also
allowed avatars to stand on volume detect objects and to have some jiggling
when they were in the volume detect object. This commit fixes both.
don't happen until the object is completely initialized.
This fixes the problem of doing a teleport while the simulator is running.
The destruction of the physical object while the engine is running means
that the physics parameter update would overwrite the new position of the
newly created avatar.
Used to detect if a simulator is receiving significant junk UDP
Decimates the number of packets between which a warning is logged and prints the IP source of the last malformed packet when logging
IncomingCloseAgent() now sets the scene presence state machine properly, which is necessary to avoid races between multiple sources of close.
Hence, it's also necessary for everyone to consistently call IncomingCloseAgent()
Calling RemoveClient() directly is currently generating an attention-grabbing exception though this right now this is harmless.
taint time. Attempt to fix a problem of teleporting within the same
region where the remove and addition of the physical avatar occasionally
ends up with a non-moving avatar.
is not with a volume detect object.
This fixes a problem of avatars trying to step over a volume detect object
that they collide with. This appeared as the avatar popping up as it started
to step up but then continuing on since the object wasn't physically interacting.
I think there is an argument for sending this information to NPCs anyway since in some cases it appears a lot easier to write server-side bots by hooking into such internal events.
However, would need to stop event messages building up on NPC queues if they are never retrieved.
This is giving much better results on teleports between simulators over my lan where for some reason there is a pause before the receiving simulator processes UpdateAgent()
At this point, v2 teleports between neighbour and non-neighbour regions on a single simulator and between v2 simulators and between a v1 and v2 simulator
are working okay for me in different scenarios (e.g. simple teleport, teleport back to original quickly and re-teleport, teleport back to neighbour and re-teleport. etc.)
This was occuring because teleport to B did not set DoNotCloseAfterTeleport on A as it was a neighbour (where it isn't set to avoid the issue where the source region doesn't send Close() to regions that are still neighbours (hence not resetting DoNotCloseAfterTeleport).
Fix here is to still set DoNotCloseAfterTeleport if scene presence is still registered as in transit from A
The root cause was that v2 was only closing neighbour agents if the root connection also needed a close.
However, fixing this requires the neighbour regions also detect when they should not close due to re-teleports re-establishing the child connection.
This involves restructuring the code to introduce a scene presence state machine that can serialize the different add and remove client calls that are now possible with the late close of the
This commit appears to fix these issues and improve teleport, but still has holes on at least quick reteleporting (and possibly occasionally on ordinary teleports).
Also, has not been completely tested yet in scenarios where regions are running on different simulators
This fixes the problem of avatars bouncing when logged in.
Added a little height to the avatar height fudges to eliminate a problem
of feet being in the ground a bit.
For unknown reasons, a dynamic function signature cannot have more than 5
parameters. Error message now tells you this fact so you can curse MS and
then go change your function definitions.
Disabled by default. Enable by setting
[Startup]ManagedStatsRemoteFetchURI="Something"
and thereafter "http://ServerHTTPPort/Something/" will return all the managed
stats (equivilent to "show stats all" console command).
Accepts queries "cat=", "cont=" and "stat=" to specify statistic category,
container and statistic names. The special name "all" is the default and returns
all values in that group.
This stops OpenSimulator still trying to teleport the user if they hit cancel on the teleport screen or closed the viewer whilst the protocol was trying to create an agent on the remote region.
Ideally, the code may also attempt to tell the destination simulator that the agent should be removed (accounting for issues where the destination was not responding in the first place, etc.)
This can currently only be activated with the console command "debug stats record start".
Off by default.
Records to file OpenSimStats.log for simulator and RobustStats.log for ROBUST
This is for testing and debugging purposes to help determine whether a particular issue may be teleport related or not
"SIMULATION/0.2" (the newer teleport protocol) remains the default. If the source simulator only implements "SIMULATION/0.1" this will correctly allow fallback to the older protocol.
Specifying "SIMULATION/0.1" will force the older, less efficient protocol to always be used.
scripts as boxed integers rather than proper reference to a new LSLInteger.
This fixes an exception when using a registered integer constant in
a script.
This is an experimental setting to control cpu spikes when an attachment heavy avatar logs in or avatars with medium attachments lgoin simultaneously.
It inserts a ms sleep specified in terms of attachments prims after each rez when an avatar logs in.
Default is 0 (no throttling).
"debug attachments <level>" changes to "debug attachments log <level>" which controls logging. A logging level of 1 will show the throttling performed if applicable.
Also adds "debug attachments status" command to show current throttle and debug logging levels.
What I believe is happening is that on initial terrain send, this is done one packet at a time.
With WaitOne, the outbound loop has enough time to loop and wait again after the first packet before the second, leading to a slower send.
This approach instead does not wait if a packet was just sent but instead loops again, which appears to lead to a quicker send without losing the cpu benefit of not continually looping when there is no outbound data.
cap is something other than "localhost". A new interface for handling
external caps is supported with an example implemented for Simian. The
only linden cap supporting this interface right now is the GetTexture
cap.
service access capabilities. In conjunction with the corresponding Simian
updates, this enables explicit per-simulator capability-based access to
grid services. That enables grid owners to add or revoke access to the grid
on a simulator by simulator basis.
Only HG Visitors get this var, to avoid spamming local users.
The config var is now called MapTileURL, to be consistent with the login one, and its being picked up from either [LoginService], [HGWorldMap] or [SimulatorFeatures], just because I have a bad memory.
This reflects the actual use of this stat - it hasn't recorded general exceptions for some time.
Make the sim extra stats collector draw the data from the stats manager rather than maintaing this data itself.
The major departure from flotsam is to send only one message per destination region, as opposed to one message per group member. This reduces messaging considerably in large groups that have clusters of members in certain regions.
This is a test setup failure since code paths when adding a duplicate root scene presence now require the EntityTransferModule to be present.
Test fixed by adding this module to test setup
These were genuine failures caused by ScenePresence.CompleteMovement() waiting for an UpdateAgent from NPC introduction that would never come.
Instead, we do not wait if the agent is an NPC.
- Child and root agents are only closed after 15 sec, maybe
- If the user comes back, they aren't closed, and everything is reused
- On the receiving side, clients and scene presences are reused if they already exist
- Caps are always recreated (this is where I spent most of my time!). It turns out that, because the agents carry the seeds around, the seed gets the same URL, except for the root agent coming back to a far away region, which gets a new seed (because we don't know what was its seed in the departing region, and we can't send it back to the client when the agent returns there).
This is because a returning viewer by teleport before 15 seconds are up will be disrupted by the close.
The 2 second delay is within the scope where a normal viewer would not allow a teleport back anyway.
Simulator/0.2 (V2) protocol will continue with the longer delay since this is actually the behaviour viewers get from the ll grid
and an early close causes other issues (avatar being sent to infinite locations temporarily, etc.)
This prevents an issue if the user teleports back to the neighbour simulator of a source before 15 seconds have elapsed.
This more closely emulates observed linden behaviour, though the timeout there is 50 secs and applies to all the pre-teleport agents.
Currently sticks a DoNotClose flag on ScenePresence though this may be temporary as possibly it could be incorporated into the ETM state machine
In this new protocol, and as committed before, the viewer is not sent EnableSimulator/EstablishChildCommunication for the destination. Instead, it is sent TeleportFinish directly. TeleportFinish, in turn, makes the viewer send a UserCircuitCode packet followed by CompleteMovementIntoRegion packet. These 2 packets tend to occur one after the other almost immediately to the point that when CMIR arrives the client is not even connected yet and that packet is ignored (there might have been some race conditions here before); then the viewer sends CMIR again within 5-8 secs. But the delay between them may be higher in busier regions, which may lead to race conditions.
This commit improves the process so there are are no race conditions at the destination. CompleteMovement (triggered by the viewer) waits until Update has been sent from the origin. Update, in turn, waits until there is a *root* scene presence -- so making sure CompleteMovement has run MakeRoot. In other words, there are two threadlets at the destination, one from the viewer and one from the origin region, waiting for each other to do the right thing. That makes it safe to close the agent at the origin upon return of the Update call without having to wait for callback, because we are absolutely sure that the viewer knows it is in th new region.
Note also that in the V1 protocol, the destination was getting UseCircuitCode from the viewer twice -- once on EstablishAgentCommunication and then again on TeleportFinish. The second UCC was being ignored, but it shows how we were not following the expected steps...
If this consistently increases then this is a problem since it means the simulator is receiving more requests than it can distribute to other parts of the code.
If we're not receiving packets with multiple threads (m_asyncPacketHandling) then this is critical since it will limit the number of incoming UDP requests that the region can handle and affects packet loss.
If m_asyncPacketHandling then this is less critical though a long process will increase the scope for threads to race.
This is an experimental stat which may be changed.
AgentUpdate packet. This fixes the problem with vehicles not moving forward
after the first up-arrow.
Code to fix a potential exception when using different IClientAPIs.
into the linkset implementation classes.
Add HasSomeCollision attribute that remembers of any component of
a linkset has a collision.
Update vehicle code (BSDynamic) to use the HasSomeCollision in place of
IsColliding to make constraint based linksets properly notice the ground.
Add linkset functions to change physical attributes of all the members
of a linkset.
- The existing event to scene has been split into 2: OnAgentUpdate and OnAgentCameraUpdate, to better reflect the two types of updates that the viewer sends. We can run one without the other, which is what happens when the avie is still but the user is camming around
- Added thresholds (as opposed to equality) to determine whether the update is significant or not. I thin these thresholds are ok, but we can play with them later
- Ignore updates of HeadRotation, which were problematic and aren't being used up stream
Added here since it was the most convenient place
Number is in the last column, "Sig. AgentUpdates" along with percentage of all AgentUpdates
Percentage largely falls over time, most cpu for processing AgentUpdates may be in UDP processing as turning this off even earlier (with "debug lludp toggle agentupdate" results in a big cpu fall
Also tidies up display.
Enabling this will stop anybody from moving on a sim, though all other updates should be unaffected.
Appears to make some cpu difference on very basic testing with a static standing avatar (though not all that much).
Need to see the results with much higher av numbers.
This appears to improve cpu usage since launching a new thread is more expensive than performing a small amount of inline logic.
However, needs testing at scale.
move around when standing on a stationary object.
Create proper linkage between BSCharacter and its actor by generating
a UpdatedProperties event the same way BSPrim does.
successfully tested, and I'm merging back those changes, which proved to
be good.
Revert "Revert "Cleared up much confusion in PollServiceRequestManager. Here's the history:""
This reverts commit fa2370b32e.
When Melanie added the web fetch inventory throttle to core, she made the long poll requests (EQs) effectively be handled on an active loop. All those requests, if they existed, were being constantly dequeued, checked for events (which most often they didn't have), and requeued again. This was an active loop thread on a 100ms cycle!
This fixes the issue. Now the inventory requests, if they aren't ready to be served, are placed directly back in the queue, but the long poll requests aren't placed there until there are events ready to be sent or timeout has been reached.
This puts the LongPollServiceWatcherThread back to 1sec cycle, as it was before.
This reverts commit 21a09ad3ad.
After more analysis and discussion, it is apparant that the Count(), Contains() and GetQueueArray() cannot be made thread-safe anyway without external locking
And this change appears to have a positive impact on performance.
I still believe that Monitor.Exit() will not release any thread for Monitor.Wait(), as per http://msdn.microsoft.com/en-gb/library/vstudio/system.threading.monitor.exit%28v=vs.100%29.aspx
so this should in theory make no difference, though mono implementation issues could possibly be coming into play.
This reverts commit 42e2a0d66e
Reverting because unfortunately this introduces race conditions because Contains(), Count() and GetQueueArray() may now end up returning the wrong result if another thread performs a simultaneous update on m_queue.
Code such as PollServiceRequestManager.Stop() relies on the count being correct otherwise a request may be lost.
Also, though some of the internal queue methods do not affect state, they are not thread-safe and could return the wrong result generating the same problem
lock() generates Monitor.Enter() and Monitor.Exit() under the covers. Monitor.Exit() does not cause Monitor.Wait() to exist, only Pulse() and PulseAll() will do this
Reverted with agreement.
This adds explicit cap poll handler supporting to the Caps classes rather than relying on callers to do the complicated coding.
Other refactoring was required to get logic into the right places to support this.
Justin Clark-Casey (justincc)