[
https://issues.apache.org/jira/browse/PHOENIX-3064?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15372797#comment-15372797
]
James Taylor commented on PHOENIX-3064:
---------------------------------------
Agreed, [~lhofhansl] - that's why I filed PHOENIX-2885 and marked it for 4.9.0
release (and implemented the UPDATE_CACHE_FREQUENCY in 4.7.0 as a partial fix).
Let's discuss on that JIRA? I believe we can handle all cases when a column,
table, view, index, sequence is referenced but doesn't exist (by updating the
cache at that point). Good point about SELECT * - that case can't be handled
100% correctly either. We also can't know when a column has been dropped (as
it'll be in our cache). Same with a view, local index, view index, and sequence
(any Phoenix-only entity). We could detect when a table or global index is
dropped (b/c the HBase metadata will be missing).
> Reduce client time spent in createTableRef and ARRQ for point queries
> ----------------------------------------------------------------------
>
> Key: PHOENIX-3064
> URL: https://issues.apache.org/jira/browse/PHOENIX-3064
> Project: Phoenix
> Issue Type: Bug
> Reporter: Lars Hofhansl
> Attachments: PhoenixGetTableAndARRQ.png
>
>
> Did some profiling for point queries (I actually wanted to test the HBase
> server side).
> I found that a _lot_ of client time is spent in
> FromCompiler.getResolverForQuery (in createTableRef) and also in
> AbstractRoundRobinQueue.offer.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
