Update Github Issue summarization codelab

[richardsliu]

This is still referencing ksonnet: https://codelabs.developers.google.com/codelabs/cloud-kubeflow-e2e-gis/index.html#5

[issue-label-bot[bot]]

Issue-Label Bot is automatically applying the label kind/bug to this issue, with a confidence of 0.60. Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback!

Links: app homepage, dashboard and code for this bot.

[sarahmaddox]

/assign @texasmichelle please would you take a look at this issue and assign it to someone else if appropriate.

[richardsliu]

/cc @texasmichelle

[richardsliu]

@amygdala Do you know who owns this codelab?

[amygdala]

Unfortunately, it no longer has an owner amongst people who work on KF. (I'm the owner of the one in the 'pipelines' subdir, but not this one). @velascoluis was just testing it and said it didn't work properly. Luis, could you add the details of how it broke, to this bug? Was it ksonnet-related?

We may need to just deprecate it for now, and possibly move the pipelines subdir one to under the 'pipelines' dir at the top level.

[amygdala]

Apparently Drew is working on a related qwiklabs replacement -- they pulled the qwiklabs version of this. (for googlers: https://buganizer.corp.google.com/issues/146655130) Drew, is there a GH issue for your new work? @dhodun

[dhodun]

We replaced the qwiklab with the current end-to-end with digit classification and did not update the github summarization code.

[richardsliu]

Should we just deprecate this codelab then?

[dhodun]

Potentially? It's pretty old at this ponit and seemed to serve the same purpose as https://www.kubeflow.org/docs/gke/gcp-e2e/

At cursory glance, only major differences were use-case and seldon core for serving vs tf serving.

[velascoluis]

Hi, following step by step the issues I found are related to ksonnet. The code itself works good inside the notebook. For googlers here is ks error ouput and the log from the execution inside the notebook: https://docs.google.com/spreadsheets/d/12eo6sc1S_tKAyBpi1iMPeoixOt4UfmMrNnImXaxWxQM/edit#gid=0

[amygdala]

For now we're in the process of 'hiding' this codelab from the codelab index, and have also marked the text 'deprecated' for those who had the link bookmarked.
We should similarly mark the readme in kf/examples -- I'll put in a PR for that shortly.

For the record, here's Luis' notes when trying to run it: ""The kfctl init fails with Error ""Error opening a gzip reader for /tmp/getter116224617/archive: EOF"", this is a reported error. I will using my v1.0 already create cluster. The install kubeflow needs to be updated part, it mentions things such as installing seldon, accesing UI via kubectl port-forward svc/ambassador 8080:80 ... I was NOT able to launch the training TFJob , everything is using ksonnet and seldon. Erro is

velascoluis@cloudshell:~/kf_examples/examples/github_issue_summarization/ks_app (velascoluis-test)$ ks generate tf-job-simple-v1beta1 tfjob --name tfjob-issue-summarization
ERROR no prototype names matched 'tf-job-simple-v1beta1'
velascoluis@cloudshell:~/kf_examples/examples/github_issue_summarization/ks_app (velascoluis-test)$ ks prototype list
NAME                                  DESCRIPTION
====                                  ===========
io.ksonnet.pkg.configMap              A simple config map with optional user-specified data
io.ksonnet.pkg.deployed-service       A deployment exposed with a service
io.ksonnet.pkg.namespace              Namespace with labels automatically populated from the name
io.ksonnet.pkg.single-port-deployment Replicates a container n times, exposes a single port
io.ksonnet.pkg.single-port-service    Service that exposes a single port 

IMHO this all should be migrated to kustomize and TFServing like in the prior example. TL;DR: This example seems pretty old and should be adapted. Does it makes sense to migrate to something like this? -> https://screenshot.googleplex.com/wDUr2DBFf38.png "

[amygdala]

The codelab has been 'deprecated' for now (& hidden in the index), but let's leave this open as a reminder to decide whether to update it or phase it out if we get more bandwidth.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.