Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Aug 20 14:38
    snikitin-me starred scalecube/scalecube-services
  • Aug 20 13:47

    io-scalecube-ci on develop

    Merge pull request #55 from sca… Merge pull request #58 from sca… Merge pull request #58 from sca… and 1 more (compare)

  • Aug 20 13:44

    io-scalecube-ci on master

    Merge pull request #58 from sca… (compare)

  • Aug 20 13:44

    io-scalecube-ci on v2.8.7

    (compare)

  • Aug 20 13:43

    io-scalecube-ci on master

    Merge pull request #58 from sca… (compare)

  • Aug 20 13:35

    artem-v on master

    ++++ Prepare for next developme… Cosmetic fix to http request-re… Merge pull request #56 from sca… and 3 more (compare)

  • Aug 20 13:35
    artem-v closed #58
  • Aug 20 13:14
    artem-v opened #58
  • Aug 20 13:14

    artem-v on update-scalecube-services-version

    (compare)

  • Aug 20 13:14

    artem-v on develop

    Updated scalecube-services.vers… Merge pull request #57 from sca… (compare)

  • Aug 20 13:14
    artem-v closed #57
  • Aug 20 13:09

    io-scalecube-ci on develop

    Merge pull request #209 from sc… Merge pull request #209 from sc… ++++ Prepare for next developme… (compare)

  • Aug 20 13:08

    artem-v on feature-update-aeron-to-latest-versions

    (compare)

  • Aug 20 13:08

    artem-v on develop

    Upgraded aeron version to latest Merge pull request #18 from sca… (compare)

  • Aug 20 13:08
    artem-v closed #18
  • Aug 20 13:07
    artem-v opened #57
  • Aug 20 13:07

    artem-v on update-scalecube-services-version

    Updated scalecube-services.vers… (compare)

  • Aug 20 13:03

    io-scalecube-ci on master

    Merge pull request #209 from sc… (compare)

  • Aug 20 13:03

    io-scalecube-ci on v2.1.16

    (compare)

  • Aug 20 13:03

    io-scalecube-ci on master

    Merge pull request #209 from sc… (compare)

Eugene Utkin
@eutkin
First of all, no one will know about this tool unless we provide it out of the box. Secondly, no one usually follows the development guidelines :) It is a rare case when someone in the project has at least check style plugin. IMHO
Good developers won't need this tool and bad developers won't use any additional tools))
Eugene Utkin
@eutkin
In the distant future, I would like to replace jdkProxy for serviceCall with AOP :)
Ronen
@ronenhamias
i think good developers master their IDE and use anything that makes them productive
but anyway
i think it should be optional at first
if we see it really helps
and provide real value then it can be mandatory or by default
Artem Vysochyn
@artem-v

no one usually follows the development guidelines

Exactly, no-one is not following. But when no-one joins some community it becomes someone and must follow development guidelines (and install checkstyle).

Ronen
@ronenhamias
well
first we need to understand why something brings value
and then see the proper solution for this
Ronen
@ronenhamias
@eutkin how about taking some issues we already have?
we have already some great ideas how to improve the project.
Ronen
@ronenhamias
for example on of the things we are really lacking is
scalecube/scalecube-services#525
Eugene Utkin
@eutkin
@artem-v This isn't about community. We are talking about projects that use scalecube service as a dependency
Eugene Utkin
@eutkin
@ronenhamias In my opinion, the documentation should be written first, and the formatting method should not be a problem. Even the current documentation is outdated. Once I've finished integrating with spring boot, I can join in writing it (.md or .aciidoc)
It would be great to see the roadmap project to have an idea of where it's going.
Ronen
@ronenhamias
i think its better to check how its going to be written
and put the correct infra
so the writting will not be done twice
as we already have one
but not so easy to write in it
and not connected to code base
so i guess we need both
and this issue will help alot
its not matter how exactlly we will have doc as long its there and maintained
Swapnil Pandey
@swapnil-pandey
Is anyone using scalecube-cluster gossip/unicast inside docker containers? I'm having a really hard time getting it to work. If anyone has done this before I would surely want to know how.
Ronen
@ronenhamias
Sure we are using it over dockers and Kubernetes
here is example
https://github.com/scalecube/scalecube-configuration-service/blob/develop/scalecube-configuration/Dockerfile
@swapnil-pandey
Swapnil Pandey
@swapnil-pandey
@ronenhamias actually I've a spring project where I have included scalecube's pom. There are multiple nodes. The project runs fine directly on the host, but when I try to dockerize the app and run it, the gossip fails to happen. It must be something to do with the docker swarm/overlay networks. But I'm failing to figure out the correct way to run the app in swarm/overlay configuration.
Artem Vysochyn
@artem-v
@swapnil-pandey try to specify memberHost and memberPort in the configuration.
Artem Vysochyn
@artem-v
@swapnil-pandey downgrade log level to trace
Ronen
@ronenhamias
@swapnil-pandey you must make sure you have a network configured and dockers have access to allocated ports.
try to specify the host and port in configuration and or make sure on what ports they are trying to connect and that they can connect.
Swapnil Pandey
@swapnil-pandey
@artem-v @ronenhamias thanks a lot for your help. The issue is now resolved. The problem was related to ports.
Ronen
@ronenhamias
@swapnil-pandey thanks for the feedback.
Swapnil Pandey
@swapnil-pandey
Is there any way to discard a packet on the receiver side which has originated say t minuted ago?
Actually, I'm using scalecube-cluster for gossiping as well as unicast sending of messages. The messages are being sent say every 1 second to the peers. But often due to network latency or otherwise, peers receive older packets from the sender which by then have become meaningless. Is there a way to ensure that I get only the latest packets?
Ronen
@ronenhamias
you cant filter it ?
there is some time to live keeping them
there is time to spead and infection period
Ronen
@ronenhamias
but i am not sure its what you try to achive...
@swapnil-pandey
Swapnil Pandey
@swapnil-pandey
How do the messages received get stored? In a queue? or something else? Can't find some docs where I could get some information about the internals.
Ronen
@ronenhamias
the message is kept across all nodes in memory
the best way is to read the class GossipProtocolImpl
and the doc
http://www.cs.cornell.edu/projects/Quicksilver/public_pdfs/SWIM.pdf
since cluster implements this one to one
@swapnil-pandey