Halt/up doesn't restart Druid

[risdenk]

I ran into this when trying to release Apache Calcite 1.19.0 and would stop the VM between RCs. When doing vagrant up, Druid wouldn't work. I needed to destroy and then recreate to get druid to start again.

[risdenk]

From @valtroffuture on the Apache Calcite dev list:

Executing steps described at these links
https://calcite.apache.org/docs/howto.html#running-integration-tests
https://github.com/vlsi/calcite-test-dataset

shows this error when accessing Druid data:
curl: (7) Failed to connect to localhost port 8082: Connection refused

Using "vagrant ssh" command and then "ps aux | grep
org.apache.druid.cli.Main" shows that Druid is not running. Farhter
research into the VM shows that it has scripts to isntall and run Druid
(install.sh, run.sh) but they all fail:
run.sh fails with the "No such file or directory" error and "Error: Could
not find or load main class org.apache.druid.cli.Main" error

The articles make it sound like Druid should be running after you execute
the "vagrant up" command.

---

Same with following steps described at
https://github.com/vlsi/calcite-test-dataset - running the "gfsh" command
shows "gfsh: command not found" error

[valtroffuture]

This issue reproes for me as well: "When doing vagrant up, Druid wouldn't work. I needed to destroy and then recreate to get druid to start again.".

[jszeluga]

I was able to successfully start Druid without destroying the vm. Once the machine starts do "vagrant ssh" to get into the machine and then become root "sudo su". Go to root's home folder (cd ~) and in there is the apache-druid folder and the zookeeper folder. Enter into the zookeeper folder and run "./bin/zkServer.sh start". Then go to the druid folder and run "run.sh". Now it should start.
So here are the steps:

1. vagrant up
2. vagrant ssh
3. sudo su
4. cd ~
5. cd zookeeper-3.4.11
6. ./bin/zkServer.sh start
7. cd ../apache-druid-0.14.0-incubating
8. ./run.sh

A startup script should probably be added