问题
I would like to create a high availability ckan in Azure. I know I can have a postgres as a service and rediss as a service, I can have ckan run behind a load balancer, only thing that bothers me is solr. Since there is no Solr as a service in Azure, how can I make sure my solr instance is highly available.
Does ckan keeps data in postgres and a duplicate of that data in solr for indexing? If yes then, that will help me as if anytime my solr instance goes does in a VM (rarely) I will create new one and replicate the data from postgres.
回答1:
Firstly, you can have SOLR with HA configuration using the SOLR Zookeeper
https://lucene.apache.org/solr/guide/6_6/setting-up-an-external-zookeeper-ensemble.html#SettingUpanExternalZooKeeperEnsemble-SettingupaZooKeeperEnsemble
Secondly, you can always reindex the SOLR in CKAN using the following command:-
https://docs.ckan.org/en/ckan-2.7.3/maintaining/paster.html#search-index-rebuild-search-index
来源：https://stackoverflow.com/questions/58568270/is-data-in-solr-backed-by-postgres-in-ckan