Last tested: Apr 19, 2020
2020-04-08 10:05:54.770 +0000 [INFO|0097a|event:render] :: Render job data not found by user 12020-04-08 10:05:54.772 +0000 [ERROR|0097a|] :: Looker::Render::RendererNoDataError : Render job 9d16e8a23065a16edc63f09c21a04a51 data not found uri:classloader:/helltool-component-build/lib/helltool/render/manager.class:491:in `get_render_results' uri:classloader:/helltool-component-build/lib/helltool/controllers/render_controller.class:163:in `block in GET /render/result/:slug/:filename'
2020-04-08 10:05:55.021 +0000 [WARN|0097a|http] :: (0.306463s) Request complete: 500 /render/result/9d16e8a23065a16edc63f09c21a04a51/Internal%20Price%20Checker%202020-04-08T0304.pdf
One of the possible causes that might trigger this exception that I came across was the fact that time was out of sync on one of the new nodes that were added to the cluster. We essentially treat nodes that have a bad heartbeat as impostors. It'll still pick up tasks but it won't be allowed to talk to the other nodes using the node-to-node endpoints which is what was happening here. it was listening to a queue that any node in the cluster is allowed to listen to and then trying to talk to a cache on another node via node-to-node but that protocol was treating it as an impostor.
You can check current setting with
MacBook-Pro-2:~ Alex$ date Sun 19 Apr 2020 18:37:24 IST