https://homeproj.cesnet.cz/https://homeproj.cesnet.cz/httpauth-login/favicon.ico?16194486082018-10-19T12:54:17ZHomeproj: Redmine for CESNETMentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=154692018-10-19T12:54:17ZPavel Káchaph@cesnet.cz
<ul><li><strong>Related to</strong> <i><a class="issue tracker-2 status-5 priority-3 priority-lowest closed" href="/issues/4348">Feature #4348</a>: Better support for sparse columns</i> added</li></ul> Mentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=154732018-10-21T12:04:52ZRadko Krkoškrkos@cesnet.cz
<ul></ul><p>Tests reveal that converting empty arrays to NULLs have no impact on table size and a negligible impact on the combined GIN index (around 1MB). No performance testing was done to assess the impact of somewhat differently organized new index.</p> Mentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=154742018-10-21T12:06:21ZRadko Krkoškrkos@cesnet.cz
<ul></ul><p><a class="user active" href="https://homeproj.cesnet.cz/users/391">Radko Krkoš</a> wrote:</p>
<blockquote>
<p>around 1MB</p>
</blockquote>
<p>This was tested on mentat-dev, so the combined GIN index is about 450MB in size.</p> Mentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=156242018-10-25T15:08:34ZPavel Káchaph@cesnet.cz
<ul></ul><p>By using my crystal sphere I presume the performance impacts would be also negligible.</p>
<p>Are there any other strong pro/con arguments, save for consistency?</p> Mentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=156262018-10-25T15:23:39ZRadko Krkoškrkos@cesnet.cz
<ul></ul><p><a class="user active" href="https://homeproj.cesnet.cz/users/17">Pavel Kácha</a> wrote:</p>
<blockquote>
<p>By using my crystal sphere I presume the performance impacts would be also negligible.</p>
</blockquote>
<p>The (manual) testing has progressed. It seems there is a performance improvement. Of course not a large one (originally in 1.5s - 2s range, now 60ms - 500ms) but an interesting one I would say (and with somewhat lower overall load - fewer workers).</p>
<p><strong>Disclaimer:</strong> These are not final numbers, the impact of ENUMs must be ruled out. Nevertheless, this seems to be universal for all newly NULLed columns.</p>
<blockquote>
<p>Are there any other strong pro/con arguments, save for consistency?</p>
</blockquote>
<p>I do not see any strong ones. I fear we have collected almost all the low-hanging fruit by now (maybe except for <a class="issue tracker-2 status-5 priority-4 priority-default closed" title="Feature: Split jsonb column into its own table (Closed)" href="https://homeproj.cesnet.cz/issues/4275">#4275</a>, but that is WiP; and low-hanging read: considerable performance benefit, not ease of implementation). Expect no further silver bullets. <span class="wiking smiley smiley-smiley" title=":)"></span></p> Mentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=160252019-01-02T14:26:45ZPavel Káchaph@cesnet.cz
<ul><li><strong>Status</strong> changed from <i>New</i> to <i>Rejected</i></li></ul><p>After discussion seems like mostly futile effort. Closing.</p> Mentat - Feature #4385: Unify empty arrays and NULLshttps://homeproj.cesnet.cz/issues/4385?journal_id=168982019-02-06T16:28:36ZJan Machjan.mach@cesnet.cz
<ul><li><strong>Target version</strong> changed from <i>Backlog</i> to <i>Rejected</i></li></ul>