sandcrawler/scald-mvp/src/test/scala/example, branch master [no description] https://git.bnewbold.net/sandcrawler/atom?h=master 2018-05-24T07:02:36+00:00 rename jvm/scalding directories 2018-05-24T07:02:36+00:00 Bryan Newbold bnewbold@archive.org 2018-05-23T19:27:59+00:00 urn:sha1:4ba428db30593b67283dd90b92141f99840dc78e get quorum fields to match, fixing test 2018-05-24T07:02:36+00:00 Bryan Newbold bnewbold@archive.org 2018-05-22T23:21:51+00:00 urn:sha1:58447ff883db0331bfd2450b8607ba332f53b3b6 Writing this commit message in anger: It seems that the HBaseSource must match exactly between the instantiated Job class and the JobTest. The error when this isn't the case is very obscure: a `None.get()` exception deep in SpyGlass internals. Blech. This may or may not explain other test failure issues. more tests (failing) 2018-05-22T19:07:01+00:00 Bryan Newbold bnewbold@archive.org 2018-05-22T19:07:01+00:00 urn:sha1:a5a3df30bad2250b4f50e6234dc810efaeaceb2f spyglass/hbase test examples (from upstream) 2018-05-22T01:36:43+00:00 Bryan Newbold bnewbold@archive.org 2018-05-22T01:36:43+00:00 urn:sha1:b2c59428fec807ef4dbe4cedcec3614154334817 fix WordCountJob package; tests; hadoop version 2018-05-21T19:25:16+00:00 Bryan Newbold bnewbold@archive.org 2018-05-21T19:25:13+00:00 urn:sha1:e517b2dd05d29e36b62b3ddfff4537cb980928f0 When copying from upstream scalding, forgot to change the path/namespace of the WordCountJob. Production IA cluster is actually running Hadoop 2.5, not 2.6 (I keep forgetting). Pull in more dependencies so test runs (copied from scalding repo, only changed the namespace of the job)