solr2rabbitmq
solr2rabbitmq is a job/library that asynchronously format and publish data from Solr query to the RabbitMQ.
Installation
You can install this library easily with pip.
pip install psql2rabbitmq
Usage
As a library
import os
import asyncio
from psql2rabbitmq import run
if __name__ == '__main__':
logger = logging.getLogger("solr2rabbitmq")
logger.setLevel(os.environ.get('LOG_LEVEL', "DEBUG"))
handler = logging.StreamHandler()
handler.setFormatter(
logging.Formatter(
os.environ.get('LOG_FORMAT', "%(asctime)s [%(levelname)s] %(name)s: %(message)s")
)
)
logger.addHandler(handler)
config = {
"mq_host": os.environ.get('MQ_HOST'),
"mq_port": int(os.environ.get('MQ_PORT', '5672')),
"mq_vhost": os.environ.get('MQ_VHOST'),
"mq_user": os.environ.get('MQ_USER'),
"mq_pass": os.environ.get('MQ_PASS'),
"mq_exchange": os.environ.get('MQ_EXCHANGE'),
"mq_routing_key": os.environ.get("MQ_ROUTING_KEY"),
"solr_collection_url": os.environ.get("SOLR_COLLECTION_URL"),
"solr_fetch_size": int(os.environ.get("SOLR_FETCH_SIZE")),
"solr_indexdate_field": os.environ.get("SOLR_INDEXDATE_FIELD"),
"solr_json_query_file_path": os.environ.get("SOLR_JSON_QUERY_FILE_PATH"),
"data_template_file_path": os.environ.get("DATA_TEMPLATE_FILE_PATH"),
"last_index_date_file_path": os.environ.get("LAST_INDEX_DATE_FILE_PATH"),
"worker_pool_size": os.environ.get("WORKER_POOL_SIZE")
}
loop = asyncio.get_event_loop()
loop.run_until_complete(run(loop=loop, logger=logger, config=config))
This library uses aio_pika, aiohttp and jinja2 packages.
Standalone
You can also call this library as standalone job command. Just set required environment variables and run psql2rabbitmq
. This usecase perfectly fits when you need run it on cronjobs or kubernetes jobs.
Required environment variables:
- MQ_HOST
- MQ_PORT (optional)
- MQ_VHOST
- MQ_USER
- MQ_PASS
- MQ_DATA_EXCHANGE (Exchange for record publishing)
- MQ_DATA_ROUTING_KEY (Routing key for record publishing records)
- MQ_PAGINATION_EXCHANGE (Exchange for solr offset publishing. Optional when running in consumer mode)
- MQ_PAGINATION_ROUTING_KEY (Routing key for solr offset publishing. Optional when running in consumer mode)
- MQ_PAGINATION_QUEUE (Queue name for pagination offsets)
- MQ_QUEUE_DURABLE (optional, default value: True)
- SOLR_COLLECTION_URL (ex:
http://solr.local:8983/solr/publication/select
) - SOLR_FETCH_SIZE (optional, default value: 20)
- SOLR_INDEXDATE_FIELD (field that stored last index datetime)
- SOLR_JSON_QUERY_FILE_PATH (File path contain solr query json Ex:
/home/user/solr_query.json
) - DATA_TEMPLATE_FILE_PATH (File path contain reqested data template. Ex:
/home/user/template.tpl
) - LAST_INDEX_DATE_FILE_PATH (File path for storing last indexed date. Ex:
/home/user/last_indexed_date.txt
) - WORKER_POOL_SIZE (optional, default value: 10)
- LOG_LEVEL (Logging level. See: Python logging module docs)
- MODE (Mode selection for scaling. See. Scalability section.)
Example Kubernetes job:
You can see it to kube.yaml
Scalability
This job can be scalable using multiple instances as of version 1.1.0.
If you are going to run a single instance, you don't need to set the MODE environment variable (or set it "DEFAULT").
Otherwise, you need to set the MODE environment variable to PAGINATOR or DEFAULT for one instance and CONSUMER for others.
PAGINATOR mode sends the offset values for the given query to MQ so that instances running in CONSUMER mode can work
independently on the same query. In DEFAULT mode, first PAGINATOR and then CONSUMER operations run.