ccr promote

This script takes Elasticsearch support diagnostics to generate a sequence of APIs to be executed to promote CCR followers to regular indices.

If your ccr cluster contains follower from several different leader cluster, you can add the flag -l|--leader <leader_cluster> to specify which leader cluster you wish to stop following.

1. Get autofollow patterns from _ccr/auto_follow
2. Get CCR follower indices from _ccr/stats
3. Find CCR follower data streams from 2.
4. Print instruction:

• Step1. Pause auto_follow patterns
• Step2. Promote data streams
• Step3. Promote indices (pause, close, unfollow, open)
• Additionally, it prints all indices promotes (for reference)

Prerequisites

• Python3 (tested with Python version 3.x):
  • how to check Python version: python3 --version
• packaging
  • Install dependices via pip3 install packaging
• Elasticsearch support diagnostics generated for your cluster. See https://github.com/elastic/support-diagnostics for more details.

Script usage

usage: ccr_promote.py [-h] (-d diagnostics | -f FOLLOWER) [-l LEADER] [--execute]

Build required APIs to promote ccr followers.

optional arguments:
  -h, --help      show this help message and exit
  -d diagnostics  path to the unzipped Elasticsearch support diagnostics bundle
  -f FOLLOWER     specify follower cluster endpoint, ie https://es_endpoint:9200
  -l LEADER       specify name of the remote cluster (leader) currently experience downtime. If not specified, it will operate on follower indices from all remote clusters.
  --execute       Without this flag, instructions will be printed to a file. You need to specify this flag to execute commands directly to the follower cluster when -f is used.

Running the script

• To get script usage:

python3 ccr_promote.py -h

• To Run with a remote cluster:

python3 ccr_promote.py -f [https://es_endpoint:9200]
python3 ccr_promote.py -f [https://es_endpoint:9200] --execute
python3 ccr_promote.py -f [https://es_endpoint:9200] -l [leader_cluster]
python3 ccr_promote.py -f [https://es_endpoint:9200] -l [leader_cluster] --execute

• To Run with a diagnostics bundle:

python3 ccr_promote.py [path_to_diag/api-diagnostics-ccr_promote_test0]
python3 ccr_promote.py [path_to_diag/api-diagnostics-ccr_promote_test0] -l [leader_cluster]

• Diagnostics (Input) can be generated from https://github.com/elastic/support-diagnostics
  • Several input files are available in the examples folder.
• Output files are written to the diganostics folder "promote-<cluster_id>.txt"
• Logging files are written to "promote_api_builder.log"