|
| 1 | +#!/usr/bin/env python3 |
| 2 | + |
| 3 | +# Usage: |
| 4 | +# |
| 5 | +# Say you have a dump file of the _schemas topic, taken via |
| 6 | +# 'rpk topic consume _schemas' and you want to restore it |
| 7 | +# directly to the _schemas topic again. |
| 8 | +# |
| 9 | +# Run the following command that parses the dump file and |
| 10 | +# generate runnable commands. This will fail without generating |
| 11 | +# the commands when a seq is higher than the corresponding offset. |
| 12 | +# |
| 13 | +# $ ./schemas_command_generator.py schemas-dump.txt > run.sh |
| 14 | +# |
| 15 | +# Here's how to restore the records: |
| 16 | +# |
| 17 | +# 1. Delete the existing _schemas topic |
| 18 | +# 2. Restart Redpanda |
| 19 | +# 3. Create the topic, configure the '-r' option based on the cluster size: |
| 20 | +# $ rpk topic create _schemas -r 3 -c cleanup.policy=compact -c compression.type=none |
| 21 | +# 4. Run the generated script above |
| 22 | +# $ chmod 755 run.sh |
| 23 | +# $ sh run.sh |
| 24 | +# |
| 25 | +# Example outputs: |
| 26 | +# $ sh run.sh |
| 27 | +# Produced to partition 0 at offset 0 with timestamp 1662516019353. |
| 28 | +# Produced to partition 0 at offset 1 with timestamp 1662516020736. |
| 29 | +# Produced to partition 0 at offset 2 with timestamp 1662516021788. |
| 30 | +# Produced to partition 0 at offset 3 with timestamp 1662516022826. |
| 31 | +# Done |
| 32 | +# |
| 33 | + |
| 34 | +import json |
| 35 | +from pprint import pprint |
| 36 | +import re |
| 37 | + |
| 38 | + |
| 39 | +def main(): |
| 40 | + import argparse |
| 41 | + |
| 42 | + def generate_options(): |
| 43 | + parser = argparse.ArgumentParser( |
| 44 | + description='Redpanda Schema Registry _schemas Command Generator') |
| 45 | + parser.add_argument('path', type=str, help='Path to the file') |
| 46 | + return parser |
| 47 | + |
| 48 | + parser = generate_options() |
| 49 | + options, _ = parser.parse_known_args() |
| 50 | + |
| 51 | + # Formatting as a single json |
| 52 | + tmp = '[' |
| 53 | + with open(options.path) as f: |
| 54 | + for l in f.read().splitlines(): |
| 55 | + if l.startswith('}'): |
| 56 | + tmp += ('},') |
| 57 | + else: |
| 58 | + tmp += (l) |
| 59 | + tmp = re.sub('},$', '}]', tmp) |
| 60 | + j = json.loads(tmp) |
| 61 | + |
| 62 | + # Generating rpk topic create commands |
| 63 | + cmd_all = '' |
| 64 | + cmd_all += f"#!/bin/bash\n\n" |
| 65 | + for c, i in enumerate(j): |
| 66 | + d = json.loads(i['key']) |
| 67 | + if d['seq'] > i['offset']: |
| 68 | + cmd_all = f"The seq {d['seq']} is unexpectedly higher than the \ |
| 69 | +offset {i['offset']} at key {i['key']}. \nThat is it's likely broken, hence exiting...." |
| 70 | + |
| 71 | + break |
| 72 | + cmd_all += f"echo '{i['value']}' | rpk topic produce _schemas --compression none -k '{i['key']}'\n" |
| 73 | + if c == len(j) - 1: |
| 74 | + cmd_all += "echo Done" |
| 75 | + else: |
| 76 | + cmd_all += "sleep 1s\n" |
| 77 | + |
| 78 | + print(cmd_all) |
| 79 | + |
| 80 | + |
| 81 | +if __name__ == '__main__': |
| 82 | + main() |
0 commit comments