Node to url_alias table is fast with this query #787464: Boost trying to crawl Unpublished nodes

SELECT *
FROM url_alias
LEFT JOIN node ON node.nid = CAST( substring(url_alias.src, 6) AS UNSIGNED )

Need to get creative or change/add fields to the boost_cache table.

Comments

mikeytown2’s picture

mikeytown2 CreditAttribution: mikeytown2 commented

Join url_alias table to nodes in the boost_cache table

SELECT *
FROM url_alias
LEFT JOIN boost_cache ON boost_cache.page_id = substring( url_alias.src, 6 )
WHERE boost_cache.page_callback = 'node'
OR boost_cache.page_callback IS NULL
mikeytown2’s picture

mikeytown2 CreditAttribution: mikeytown2 commented

Would like to join menu_router, url_alias & boost_cache tables together with the option of joining the node, term and user tables. From what I can tell the thing that is holding me back is the url_alias table. I don't have a reliable way to join to that ATM. Means I need to create a new field or rework what I got. Seems like the best bet is the src field on the url_alias table. dst and url would not be fast due to url being a text field in the boost table & that it contains the query string.

Using the substring trick, I can do fast joins but that is not a generalized solution. concat might be an option... nope that is SLOW.