[
https://issues.apache.org/jira/browse/SOLR-6349?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man updated SOLR-6349:
---------------------------
Attachment: SOLR-6349.patch
Two steps forward, one step back: I fixed the bugs i mentioned before dealing
with localparam stats that are all 'false', and cleaned up a bit of the stats
returned/calculated logic -- this lead me to understanding that the "whitebox"
isShard test i was confused by before was definitely broken (see below) ... i
tried to use SOLR-7147 to do a true whitebox test and inspect _real_ shard
requests/responses, but got blocked by some bugs in the test framework
Full notes about changes in this patch...
* StatsField
** add Set<Stat> DEFAULT_STATS
** populateStatsRequested(...)
*** rename to populateStatsSets() and eliminate args -- method is already
mucking with other state directly
*** use DEFAULT_STATS if and only if there are no stat localparams (even if
false)
** calculateStats()
*** fix to not assume anything about the empty set - empty set means calculate
nothing
* StatsComponentTest
** testIndividualStatLocalParams
*** my review & bug fixing in StatsField confirmed that the whitebox testing
here wasn bogus - the types of "isShard=true" requests simulated here don't
actaully look like what a real shard request looks like. This test _was_
asking for each of the individual stats that appear as in getDistribDeps() of
any stat in an isShard request and confirmed that only that single stat was
returned -- but the actual code doesn't ask for individual distribDeps that
way. it asks for the oroginally requested stat and trusts that the isShard
indication will cause only the deps to be returned
*** fixed the whitebox queries to simulate the real requests for every stats,
and assert all deps come back - even if the original request excluded those
deps
* TestDistributedSearch
** beefed up testing of distrib requests where the deps of a stat requested by
the client are also explicitly excluded by the client (ie: "mean=true sum=false
count=false"
** started down the path of leveraging SOLR-7147 to inspect the shard req/rsp
to ensure we only get what we need - but got blocked by SOLR-7171 (see
nocommits)
> LocalParams for enabling/disabling individual stats
> ---------------------------------------------------
>
> Key: SOLR-6349
> URL: https://issues.apache.org/jira/browse/SOLR-6349
> Project: Solr
> Issue Type: Sub-task
> Reporter: Hoss Man
> Attachments: SOLR-6349-tflobbe.patch, SOLR-6349-tflobbe.patch,
> SOLR-6349-tflobbe.patch, SOLR-6349-xu.patch, SOLR-6349-xu.patch,
> SOLR-6349-xu.patch, SOLR-6349-xu.patch, SOLR-6349.patch, SOLR-6349.patch,
> SOLR-6349.patch, SOLR-6349.patch, SOLR-6349.patch,
> SOLR-6349___bad_idea_broken.patch, make-data-and-queries.pl
>
>
> Stats component currently computes all stats (except for one) every time
> because they are relatively cheap, and in some cases dependent on eachother
> for distrib computation -- but if we start layering stats on other things it
> becomes unnecessarily expensive to compute all the stats when they just want
> the "sum" (and it will definitely become excessively verbose in the
> responses).
> The plan here is to use local params to make this configurable. All of the
> existing stat options could be modeled as a simple boolean param, but future
> params (like percentiles) might take in a more complex param value...
> Example:
> {noformat}
> stats.field={!min=true max=true percentiles='99,99.999'}price
> stats.field={!mean=true}weight
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
