Comments (6)
got a measurment on the releaseChunk just now, equally bad...
Function = iox::popo::SubscriberPortUser::releaseChunk(iox::mepoo::ChunkHeader const*) [1110310]
nsecs : count distribution
0 -> 1 : 0 | |
2 -> 3 : 0 | |
4 -> 7 : 0 | |
8 -> 15 : 0 | |
16 -> 31 : 0 | |
32 -> 63 : 0 | |
64 -> 127 : 0 | |
128 -> 255 : 0 | |
256 -> 511 : 0 | |
512 -> 1023 : 0 | |
1024 -> 2047 : 0 | |
2048 -> 4095 : 0 | |
4096 -> 8191 : 14368 |****************************************|
8192 -> 16383 : 1043 |** |
16384 -> 32767 : 173 | |
32768 -> 65535 : 5 | |
65536 -> 131071 : 1 | |
131072 -> 262143 : 1 | |
from iceoryx.
maybe related to #2221 ? -Because slow consumers (with the DISCARD_OLDEST_DATA setting on publisher & subscriber, and subscriber queueCapacity=256) can't ever trigger this right ?
from iceoryx.
I don't think it is related to #2221 since it also effects the allocation not only the release of the chunks.
Do the publisher allocate from the same mempool? If this is the case, they compete for the same chunk and there is also high contention which slow down each other. You could try to create multiple mempools with slightly different chunk sizes and let each core allocate from a different mempool.
Do you also use the WaitSet/Listener? This might introduce additional latencies due to context switches.
How many subscriber does a publisher have?
Do you have multiple publisher publishing to the same topic?
from iceoryx.
Oh, can you also check with dmesg whether the TSC clock is deactivated?
It would also be helpful to check with https://github.com/KDAB/hotspot where the bottleneck lies.
It helped us to indemnify why my new shiny AMD Ryzen laptop was slower with iceoryx2 than my 6 year old Intel Laptop. It seems on some AMD CPUs the TSC clock is deactivated which results in a context switch each time clock_gettime
was used.
from iceoryx.
Thanks for the reply.
The publishers allocate the same two sized chunks yes (size2, size3). It's a bit unpractical to add padding just to facilitate mempool partitioning though. If contention can lead to significant overhead I guess the framework should maybe provide support for mempool partitioning (by id or name) ? (On a related issue we'll jack in a blocking notification callback and issue a PR.)
Yes using the 4 listeners on separate cores in this case receiving high frequency publications (size1). Will try investigate the overhead of that using hotspot...
Half of the publishers per core/listener have one subscriber (size3) the other half three subscribers (size2).
There's one publisher publishing on a shared topic (alerts) but it does so very infrequent and not during the above measurements.
The TSC clock is active on our EPYC CPU.
from iceoryx.
The issue with the contention is fixed in iceoryx2. There, each publisher has it's own data segment. With iceoryx2 we try to fix all the issues which user encounter when iceoryx is pushed to its limits.
from iceoryx.
Related Issues (20)
- Update links to default branch
- Make iceoryx resource prefix a compile time option
- Fix new clang-tidy-18 warnings
- ICEORYX error! POPO__CHUNK_LOCKING_ERROR HOT 4
- Mark iox container operations which return bool as [[nodiscard]] HOT 1
- 'iox::string' tests can exceed the translation unit compilation timeout
- Prepare v2.0.6 release for ROS2
- Compiler warnings on latest macOS on release_2.0
- While setting the acquired shared memory to zero a fatal SIGBUS signal appeared caused by memset. HOT 5
- how to support string using zero-copy transport HOT 1
- Add 'iox' prefix to all functions and types in the platform abstraction
- SingleProcess example crashes on QNX HOT 4
- : backtrace: HOT 2
- Is `libatomic` still required to build Iceoryx? HOT 7
- Some questions about IndexQueue. HOT 10
- Feature request: print the number of ports in the introspection client
- 32-Bit Support
- How to handle messages in parallel with popo::listener? HOT 3
- Is there a cost comparison between serialization app and roudi with protobuf or other efficient implementation HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from iceoryx.