aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/OpenSim/Region/ClientStack/Linden/UDP (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* Add stat clientstack.<scene>.IncomingPacketsOrphanedCount to record ↵Justin Clark-Casey (justincc)2013-08-141-4/+29
| | | | well-formed packets that were not initial connection packets and could not be associated with a connected viewer.
* Count any incoming packet that could not be recognized as an LLUDP packet as ↵Justin Clark-Casey (justincc)2013-08-141-21/+44
| | | | | | | a malformed packet. Record this as stat clientstack.<scene>.IncomingPacketsMalformedCount 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
* Merge branch 'TeleportWork'Justin Clark-Casey (justincc)2013-08-093-4/+10
|\
| * Fix an issue where under teleport v2 protocol, teleporting from regions in ↵Justin Clark-Casey (justincc)2013-08-083-4/+10
| | | | | | | | | | | | | | | | | | | | an line from A->B->C would not close region A when reaching C 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 makes in-world terrain editing smoother, even in MegaRegions. This ↵teravus2013-08-071-3/+30
|/ | | | change only affects the editing user's experience. Non-editing users will see nothing different from the current 'slow' result. See comments for the thought process and how the issues surrounding terrain editing, cache, bandwidth, threading, terrain patch reliability and throttling were balanced.
* For LLImageManagerTests, make tests execute under synchronous fire and ↵Justin Clark-Casey (justincc)2013-08-051-8/+12
| | | | | | forget conditions. I generally prefer this approach for regression tests because of the complexity of accounting for different threading conditions.
* Fix the failing TestSendImage. J2K decoding is async.Diva Canto2013-08-031-1/+9
|
* minor: Add name to debug lludp packet level feedback on consoleJustin Clark-Casey (justincc)2013-08-011-1/+1
|
* Try a different approach to slow terrain update by always cycling the loop ↵Justin Clark-Casey (justincc)2013-08-011-1/+2
| | | | | | | | immediately if any data was sent, rather than waiting. 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.
* Revert "Issue: painfully slow terrain loading. The cause is commit ↵Justin Clark-Casey (justincc)2013-08-011-5/+5
| | | | | | d9d995914c5fba00d4ccaf66b899384c8ea3d5eb (r/23185) -- the WaitOne on the UDPServer. Putting it back to how it was done solves the issue. But this may impact CPU usage, so I'm pushing it to test if it does." This reverts commit 59b461ac0eaae1cc34bb82431106fdf0476037f3.
* Issue: painfully slow terrain loading. The cause is commit ↵Diva Canto2013-08-011-5/+5
| | | | d9d995914c5fba00d4ccaf66b899384c8ea3d5eb (r/23185) -- the WaitOne on the UDPServer. Putting it back to how it was done solves the issue. But this may impact CPU usage, so I'm pushing it to test if it does.
* minor: Add timeout secs to connection timeout message. Change message to ↵Justin Clark-Casey (justincc)2013-07-291-8/+9
| | | | reflect it is a timeout due to no data received rather than an ack issue.
* Fix issue just introduced in 8efe4bfc2ed7086e9fdf4812297e6525f955f6ac where ↵Justin Clark-Casey (justincc)2013-07-291-1/+1
| | | | I accidentally left in a test line to force very quick client unack
* Make "abnormal thread terminations" into "ClientLogoutsDueToNoReceives" and ↵Justin Clark-Casey (justincc)2013-07-291-5/+21
| | | | | | | add this to the StatsManager 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.
* Deleted all [ZZZ] debug messages.Diva Canto2013-07-241-5/+0
|
* One more thing to test in order to let CompleteMovement go up the stack.Diva Canto2013-07-241-2/+3
|
* Minor adjustment on timings of waits.Diva Canto2013-07-241-4/+19
|
* Changed the RegionHandshake packet to the Unknown queue, so that it is sent ↵Diva Canto2013-07-241-1/+1
| | | | with high priority and hopefully gets to the client before AgentMovementComplete
* New Teleport protocol (V2), still compatible with V1 and older. (version of ↵Diva Canto2013-07-241-0/+68
| | | | | | | | the destination is being checked) 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...
* Add proper method doc and comments to m_dataPresentEvent (from d9d9959)Justin Clark-Casey (justincc)2013-07-231-2/+13
|
* Add clientstack.InboxPacketsCount stat. This records the number of packets ↵Justin Clark-Casey (justincc)2013-07-231-0/+13
| | | | | | waiting to be processed at the second stage (after initial UDP processing) 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.
* Add clientstack.OutgoingUDPSendsCount stat to show number of outbound UDP ↵Justin Clark-Casey (justincc)2013-07-232-3/+22
| | | | packets sent by a region per second
* Record raw number of UDP receives as clientstack.IncomingUDPReceivesCountJustin Clark-Casey (justincc)2013-07-232-1/+20
|
* Add AverageUDPProcessTime stat to try and get a handle on how long we're ↵Justin Clark-Casey (justincc)2013-07-232-2/+56
| | | | | | | | taking on the initial processing of a UDP packet. 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.
* Added check for user movement specification before discarding an incomingRobert Adams2013-07-222-42/+42
| | | | | | 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.
* Minor cosmetic changes.Diva Canto2013-07-211-27/+23
|
* The quaternion delta was a bit to high, now that the head rotation is out of ↵Diva Canto2013-07-211-1/+1
| | | | the equation. (head rotation was the problematic one)
* EDIT BEAMS!!! They had been missing from OpenSim since ever. Thanks to ↵Diva Canto2013-07-211-23/+4
| | | | lkalif for telling me how to route the information. The viewer effect is under the distance filter, so only avatars with cameras < 10m away see the beams.
* A couple of small optimizations over the previous commitDiva Canto2013-07-211-8/+10
|
* Manage AgentUpdates more sanely:Diva Canto2013-07-211-106/+108
| | | | | | - 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
* Fixed the stats in show client stats. Also left some comments with ↵Diva Canto2013-07-212-5/+15
| | | | observations about AgentUpdates.
* furhter shorten CheckAgentUpdateSignificance(). No real perf impact.Justin Clark-Casey (justincc)2013-07-211-7/+5
|
* Remove some pointless code in CheckAgentUpdateSignificance()Justin Clark-Casey (justincc)2013-07-211-30/+20
|
* Make the check as to whether any particular inbound AgentUpdate packet is ↵Justin Clark-Casey (justincc)2013-07-212-58/+129
| | | | | | significant much earlier in UDP processing (i.e. before we pointlessly place such packets on internal queues, etc.) Appears to have some impact on cpu but needs testing.
* Add measure of number of inbound AgentUpdates that were seen as significant ↵Justin Clark-Casey (justincc)2013-07-211-0/+3
| | | | | | | | | to "show client stats" (i.e. sent on for further processing instead of being discarded) 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.
* Hack in console command "debug lludp toggle agentupdate" to allow ↵Justin Clark-Casey (justincc)2013-07-211-0/+24
| | | | | | | | AgentUpdate in packets to be discarded at a very early stage. 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.
* minor: provide user feedback in the log for now when udp in/out bound ↵Justin Clark-Casey (justincc)2013-07-211-4/+8
| | | | threads are started/stopped
* Fix up a temporary debugging change from last commit which stopped "lludp ↵Justin Clark-Casey (justincc)2013-07-211-2/+1
| | | | stop out" from actually doing anything
* Do some simple queue empty checks in the main outgoing udp loop instead of ↵Justin Clark-Casey (justincc)2013-07-215-22/+92
| | | | | | | always performing these on a separate fired thread. 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.
* try Hacking in an AutoResetEvent to control the outgoing UDP loop instead of ↵Justin Clark-Casey (justincc)2013-07-182-2/+28
| | | | | | a continuous loop with sleeps. Does appear to have a cpu impact but may need further tweaking
* Revert "Puts RequestImage (UDP) back to asyn -- CPU spike hunt"Diva Canto2013-07-151-1/+1
| | | | This reverts commit b060ce96d93a33298b59392210af4d336e0d171b.
* Puts RequestImage (UDP) back to asyn -- CPU spike huntDiva Canto2013-07-151-1/+1
|
* Revert the revertDiva Canto2013-07-151-0/+2
| | | | | | Revert "Trying to hunt the CPU spikes recently experienced." This reverts commit ac73e702935dd4607c13aaec3095940fba7932ca.
* Trying to hunt the CPU spikes recently experienced.Diva Canto2013-07-151-2/+0
| | | | | | Revert "Comment out old inbound UDP throttling hack. This would cause the UDP" This reverts commit 38e6da5522a53c7f65eac64ae7b0af929afb1ae6.
* Moved SendInitialDataToMe to earlier in CompleteMovement. Moved ↵Diva Canto2013-07-131-1/+1
| | | | | | | | TriggerOnMakeRootAgent to the end of CompleteMovement. Justin, if you read this, there's a long story here. Some time ago you placed SendInitialDataToMe at the very beginning of client creation (in LLUDPServer). That is problematic, as we discovered relatively recently: on TPs, as soon as the client starts getting data from child agents, it starts requesting resources back *from the simulator where its root agent is*. We found this to be the problem behind meshes missing on HG TPs (because the viewer was requesting the meshes of the receiving sim from the departing grid). But this affects much more than meshes and HG TPs. It may also explain cloud avatars after a local TP: baked textures are only stored in the simulator, so if a child agent receives a UUID of a baked texture in the destination sim and requests that texture from the departing sim where the root agent is, it will fail to get that texture. Bottom line: we need to delay sending the new simulator data to the viewer until we are absolutely sure that the viewer knows that its main agent is in a new sim. Hence, moving it to CompleteMovement. Now I am trying to tune the initial rez delay that we all experience in the CC. I think that when I fixed the issue described above, I may have moved SendInitialDataToMe to much later than it should be, so now I'm moving to earlier in CompleteMovement.
* Switched UUIDNameRequest and RegionHandleRequest to Sync, because now they ↵Diva Canto2013-07-111-2/+2
| | | | are also non-blocking handlers.
* Switched TransferRequest (UDP packet handler) to sync. The permissions ↵Diva Canto2013-07-111-99/+115
| | | | checks may block, so they get a FireAndForget. Everything else is non-blocking.
* Switched RegionHandshakeReply to Sync, because it's not doing anything blocking.Diva Canto2013-07-101-1/+1
|
* EXPERIMENTAL: make RequestImage (UDP packet handler) sync instead of async. ↵Diva Canto2013-07-101-1/+1
| | | | This _shouldn't_ screw things up, given that all this does is to dump the request in a queue.
* Added show client-stats [first last] command to expose what viewers are ↵Diva Canto2013-07-102-15/+30
| | | | requesting.