Comments (6)
Yeah typically we submit the slim jar to our cluster when its used as a dependency by another project. I've only used the fat jar for running example jobs so it can act as a standalone. I haven't had any trouble changing versions with the fat jar myself yet, but yes you do have to update the relevant libraries if a mismatch of major scala versions occurs.
We've discussed adding some Maven build profiles to make this a bit easier, but there are many potential combos of Scala, Scalding, Cascading, and Hadoop and I've worked on other open project where maintaining a long list of build profiles ended up making it more brittle.
All the same, I'm not against the idea if you want to submit a pull request my team and I will take a look and would give the idea another look.
Thanks for your time and interest!
from sahale.
Hey update: we are upgrading to Scala 2.10.x and Scalding 0.12 soon and hit a few minor lib and code changes needed for it. I've tagged the current release as v0.5 and will tag the next as v0.6, as well as leave comments to help users navigate the incompatible changes Scala and Scalding make across versions.
Regardless, if you rebuild flow tracker and restart the NodeJS app, then depend on the new Flowtracker for your runs, everything should work as long as you match up your internal Hadoop/Cascading/Scala/Scalding versions with Sahale/Flowtracker accordingly. Fun times.
The release will be posted Wednesday or sooner. Thanks, sorry about the headache.
from sahale.
This change is pushed out, which makes a few POM.xml changes and some (sadly) code changes to stay compatible with newer Scala/Scalding versions. this commit can be reverted if you're still on older Scalding (or Scala 2.9.x) and the other new stuff should work fine.
I've opened another Issue to perhaps set up Maven build profiles to manage the Scala 2.9/2.10 divide better. I'm closing this one for now.
from sahale.
Hey, thanks again. We have an issue open to decide what to do about this.
One issue is there are actual code changes required as the Scalding side of
things changes slightly across the Scala 2.9 => 2.10 barrier, but we could
probably shim that in the build too. Sorry for the frustration, hope you
got it worked out in the meantime? Feel free to drop me a gist of any
relevant stack traces, POM changes you made etc. and I'll take a look.
On Thu, Feb 26, 2015 at 4:07 AM, m.orazow [email protected] wrote:
After changing scala version and submitting fat jar to cluster throws
error,
NoClassDefFoundError: scala/reflect/ClassManifestI think this happens because there is some classes still compiled with
lower scala versions.I think, ideal would be to have cross compiled jar in maven (or other)
repositories.What do you think?
I would be glad to help if there is anything I can do.
Best—
Reply to this email directly or view it on GitHub
#1.
from sahale.
Hello @initialcontext, thanks for informing.
It is so great to see improvements. Since I am busy with other projects I could not look at it yet. I will deploy it with new versions and let you know if I see any issues.
All the Best!
from sahale.
You too, thanks for the feedback! Let us know if you hit more trouble.
We moved to Scala 2.10 and upgraded Scalding internally, and the current FlowTracker lib is working for us so I believe at this point the dependencies and FlowTracker itself should be working properly in a Scala 2.10 and compatible Scalding env, but I'd love to know if it doesn't!
from sahale.
Related Issues (20)
- Move Existing Log Links to New "Links" Tab
- Include configurable mapping from RM/JT URLs to Hadoop Cluster names for display
- Correct Progress Bar with Map-Only Stages HOT 1
- Pull Out Real Input/Output Taps
- Pull Out Links to Error Logs
- Autoselect Stage for Single-Stage Jobs
- Factor out browser state from graph view HOT 1
- Remove CDH dependencies and update hadoop deps
- Flowtracker NPE
- Support pluggable username linking HOT 2
- Upgrade scalatest to 3.X
- Upgrade scalding to 0.17.X
- Consider providing multiple Maven profiles HOT 1
- Release Tags HOT 2
- Sahale should report bytes to/from GCS if that information is reported to it HOT 1
- If the FlowTracker can't initialize a connection to Sahale, it should not attempt to push final stats.
- Include glue code and example jobs targeting other Cascading DSLs HOT 1
- Shutdown hook leads to instability
- Set up end-to-end tests?
- Some failed steps report bad running time HOT 2
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 sahale.