GPII-3867 - Update load data section #361
Conversation
gcp/README.md
Outdated
|
|
||
| Let's load the data, go back to the folder _testData/myDbData_ and run: | ||
| 1. `curl -d @gpiiKeys.json -H "Content-type: application/json" -X POST http://localhost:5984/gpii/_bulk_docs` | ||
| 1. `curl -d @.json -H "Content-type: application/json" -X POST http://localhost:5984/gpii/_bulk_docs` | ||
| 1. `curl -d @gpiiKeys.json -H "Content-type: application/json" -X POST $GPII_DATASOURCE_HOSTNAME@localhost:5984/gpii/_bulk_docs` |
There was a problem hiding this comment.
When running inside rake sh, secret variables can be used directly, i.e. http://$TF_VAR_secret_couchdb_admin_username:[email protected]:5984/gpii/_bulk_docs.
Before I used the following command to interact with CouchDB from rake shell: rake sh["kubectl exec --namespace gpii couchdb-couchdb-0 -c couchdb -- curl -s http://$TF_VAR_secret_couchdb_admin_username:[email protected]:5984/gpii/_all_docs"]
There was a problem hiding this comment.
done, thanks @natarajaya
gcp/README.md
Outdated
| __Note__ that if you are going to perform this in production (prd) you should do it from the _prd_ folder and remember to use the _RAKE_REALLY_RUN_IN_PRD=true_ variable when issuing the commands against the production cluster. | ||
| 1. Open a port forwarding between the cluster's couchdb host:port and your local machine: `kubectl --namespace gpii port-forward couchdb-couchdb-0 5984 &` | ||
| 1. Get the current credentials for connecting to the couchdb from a flowmanager pod: `eval $(kubectl -n gpii exec flowmanager-[POD_ID] env | grep ^GPII_DATASOURCE_HOSTNAME | cut -f1 -d@)` |
There was a problem hiding this comment.
Please check my previous comment.
|
These are fine cleanups, but this is not what I requested. GPII/universal#769 (comment) |
|
@mrtyler I added the procedure I followed in the last 2 loads. |
amatas
changed the title
Yes. The goal of this task was to add documentation we will need to start the system from scratch -- https://issues.gpii.net/browse/GPII-2806. |
|
As we decided yesterday I'm going to merge this PR, the task is not completed as we still need the instructions about how to initialize the DB in the case of a disaster. |
Update the load of the data section in the README
Rendered version