General discussion of Lagom: https://www.lagomframework.com/ | Code of Conduct: https://www.lightbend.com/conduct | Forums: https://discuss.lagomframework.com | Commercial support from https://www.lightbend.com/
:)
Sorry, had to :D
(reference to Frankenstain Jr.)
:point_up: September 26, 2018 1:03 AM if you read-side tables were lost, just delete the offsets from the offset table and the readside processors will consume the
eventStream again. Read side tables are just a different representation fo the journal. As long as you don’t loose the journal you are fine (you could even delete the snapshots, but I’m not saying that is a good idea).
:point_up: September 26, 2018 6:34 AM We use k8s ConfigMap for environment specific config. It’s very useful, IMHO.
@humpahh_twitter to configure the Akka HTTP backend inside Play (which is what is powering Lagom under the covers), you need to prefix Akka HTTP settings with
play.server. See https://www.playframework.com/documentation/2.6.x/SettingsAkkaHttp . So, in your case, I think you should use:play.server.akka.http.server.parsing.max-chunk-size
@humpahh_twitter you should keep in mind that Lagom’s DevMode uses a custom HTTP server which may ignore the settings inplay.serverscope.
@ignasi35 Thanks. How/where should I configure the settings or override them somehow if I want them to be in effect in DevMode? Having those configs in play.server {} doesn't seem to do the trick now in DevMode, gives same exception from akka.http
@Sureshkatweet_twitter
- ConfigMap
apiVersion: v1 data: conf: | foo.bar = "value" play.http.secret.key = "play-http-secret-key-test" kind: ConfigMap metadata: name: service-config - Add ConfigMap data to a Volume
For example volume/opt/docker/conf/config/application.conf - Add include to base configuration file
include file("/opt/docker/conf/config/application.conf") - Profit! :smile:
PS: Of cource we have differente ConfigMap for every test (k8s namespace) and production environment.
:point_up: September 26, 2018 11:20 AM @humpahh_twitter it is possible to tune the dev mode server using the prefix
lagom.akka.dev-mode.config in the settings. Note that it is a prefix and it will only affect the server. So, if you tune play.server.akka.http.server.parsing.max-chunk-size I think you’d need to tune lagom.akka.dev-mode.config.akka.http.server.parsing.max-chunk-size with the same values.
@Sureshkatweet_twitter
- ConfigMap
apiVersion: v1 data: conf: | foo.bar = "value" play.http.secret.key = "play-http-secret-key-test" kind: ConfigMap metadata: name: service-config- Add ConfigMap data to a Volume
For example volume/opt/docker/conf/config/application.conf- Add include to base configuration file
include file("/opt/docker/conf/config/application.conf")- Profit! :smile:
PS: Of cource we have differente ConfigMap for every test (k8s namespace) and production environment.Thanks for quick reply @ihostage
Specifically, this started happening when I added a stubbed client. :-(
I wish there was a way to force a compiler error for missing lazy vals
I wondering something, when a service start, Is persistentEntity replay events for all entities ?
I am not sur about the behavior at start time
@ignasi35 For some reason it is still not overriding the default with lagom.akka.dev-mode.config.akka.http.server.parsing.max-chunk-size = "10m", it still gives the "limit of 1048576 bytes" error in Dev for some reason. https://pastebin.com/4vcb4JxU for what config I have in the impl for that service.
You've to do something called as schema evolution to handle changes to any DTO's that you'll have in the write side. https://www.lagomframework.com/documentation/1.4.x/scala/Serialization.html
Hmm there is a bit of an issue with
PlayJsonSerializer. It uses the runtime class to find the formatter but in many cases this could be anything. For example: Caused by: java.lang.RuntimeException: Missing play-json serializer for [scala.collection.immutable.HashMap$HashTrieMap] when trying to return an immutable.Map. Any solutions without having to catch these one by one and register them (and pray the class is public)?
The good thing is that the test driver usually catches these (bad thing is we didn’t have a test for this case and so didn’t find it)
guys, why would an -impl service module which is subscribe-only and only references
lagom-javadsl-kafka-client_${scala.binary.version} still have lagom-kafka-broker_2.12-1.5.0-M3 in the libraries? I’m bundling it all following the latest k8s how-to, lagom 1.5.0-M3, and fabric8’s docker-maven-plugin
I’m writing a service spec where I need to first call one service to create a parent entity and then use the returned ID of that entity when creating a child entity through a different service. The problem is eventual consistency. Should I wait for the parent entity to be available or how do people handle this scenario?
