<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: The task starts slowly in Qlik Replicate</title>
    <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076512#M6128</link>
    <description>&lt;P&gt;Hi&amp;nbsp;&lt;a href="https://community.qlik.com/t5/user/viewprofilepage/user-id/151740"&gt;@XiaoyongLin&lt;/a&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I see you also opened a severity 1 production down case for this. Copying my response here:&lt;/P&gt;
&lt;P&gt;&lt;SPAN&gt;By default, Replicate will run a "select *" on all tables in the task for the full load phase. Whether that results in a full table scan on the source depends on how the tables are indexed and how Oracle determines the best way to fetch the data is.&lt;/SPAN&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;SPAN&gt;You might consider using the parallel load option, see here for details:&lt;/SPAN&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;SPAN&gt;&lt;A href="https://help.qlik.com/en-US/replicate/May2023/Content/Global_Common/Content/SharedEMReplicate/Customize%20Tasks/Parallel_Load.htm" target="_blank"&gt;https://help.qlik.com/en-US/replicate/May2023/Content/Global_Common/Content/SharedEMReplicate/Customize%20Tasks/Parallel_Load.htm&lt;/A&gt;&lt;/SPAN&gt;&lt;/P&gt;
&lt;P&gt;Parallel load would also help to avoid ORA-01555 snapshot too old errors when handling large tables.&lt;/P&gt;
&lt;P&gt;&lt;BR /&gt;&lt;SPAN&gt;Thanks,&lt;/SPAN&gt;&lt;BR /&gt;&lt;SPAN&gt;Dana&lt;/SPAN&gt;&lt;/P&gt;</description>
    <pubDate>Thu, 25 May 2023 20:45:30 GMT</pubDate>
    <dc:creator>Dana_Baldwin</dc:creator>
    <dc:date>2023-05-25T20:45:30Z</dc:date>
    <item>
      <title>The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076509#M6127</link>
      <description>&lt;P&gt;I started a task and the wait time was very long，The log contents are as follows&lt;/P&gt;
&lt;P&gt;execute table description query 'select * from "SJZBQ_SS"."DZDZ_HWXX_DZFP"'&lt;/P&gt;
&lt;P&gt;The amount of data in this table is very large，Why does replicate scan the full table data？&lt;/P&gt;
&lt;P&gt;source database:oracle&lt;/P&gt;
&lt;P&gt;target database:oracle&lt;/P&gt;
&lt;P&gt;task mode:Incremental synchronization&lt;/P&gt;
&lt;P&gt;Wait for a response online&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 20:05:52 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076509#M6127</guid>
      <dc:creator>XiaoyongLin</dc:creator>
      <dc:date>2023-05-25T20:05:52Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076512#M6128</link>
      <description>&lt;P&gt;Hi&amp;nbsp;&lt;a href="https://community.qlik.com/t5/user/viewprofilepage/user-id/151740"&gt;@XiaoyongLin&lt;/a&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I see you also opened a severity 1 production down case for this. Copying my response here:&lt;/P&gt;
&lt;P&gt;&lt;SPAN&gt;By default, Replicate will run a "select *" on all tables in the task for the full load phase. Whether that results in a full table scan on the source depends on how the tables are indexed and how Oracle determines the best way to fetch the data is.&lt;/SPAN&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;SPAN&gt;You might consider using the parallel load option, see here for details:&lt;/SPAN&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;SPAN&gt;&lt;A href="https://help.qlik.com/en-US/replicate/May2023/Content/Global_Common/Content/SharedEMReplicate/Customize%20Tasks/Parallel_Load.htm" target="_blank"&gt;https://help.qlik.com/en-US/replicate/May2023/Content/Global_Common/Content/SharedEMReplicate/Customize%20Tasks/Parallel_Load.htm&lt;/A&gt;&lt;/SPAN&gt;&lt;/P&gt;
&lt;P&gt;Parallel load would also help to avoid ORA-01555 snapshot too old errors when handling large tables.&lt;/P&gt;
&lt;P&gt;&lt;BR /&gt;&lt;SPAN&gt;Thanks,&lt;/SPAN&gt;&lt;BR /&gt;&lt;SPAN&gt;Dana&lt;/SPAN&gt;&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 20:45:30 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076512#M6128</guid>
      <dc:creator>Dana_Baldwin</dc:creator>
      <dc:date>2023-05-25T20:45:30Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076513#M6129</link>
      <description>&lt;P&gt;Hi&amp;nbsp;&lt;a href="https://community.qlik.com/t5/user/viewprofilepage/user-id/151740"&gt;@XiaoyongLin&lt;/a&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Also:&amp;nbsp;&lt;SPAN&gt;Since you have mentioned "task mode:Incremental synchronization" I'm assuming it's a CDC only task and you are seeing select* from running in the DB during the initial metadata refresh step.&lt;/SPAN&gt;&lt;BR /&gt;&lt;SPAN&gt;Could you please let us know the Replicate version you are on? We had similar issue in 2022.11 SP04 or lower and was fixed in later Service pack.&lt;/SPAN&gt;&lt;/P&gt;
&lt;P&gt;Thanks,&lt;/P&gt;
&lt;P&gt;Dana&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 20:48:08 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076513#M6129</guid>
      <dc:creator>Dana_Baldwin</dc:creator>
      <dc:date>2023-05-25T20:48:08Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076521#M6132</link>
      <description>&lt;P&gt;Hi &lt;a href="https://community.qlik.com/t5/user/viewprofilepage/user-id/121014"&gt;@Dana_Baldwin&lt;/a&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;My Replicate version is 2022.5.0.652,This is a production task, I can't do an upgrade now, How can I solve this problem without upgrading?&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 21:18:06 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076521#M6132</guid>
      <dc:creator>XiaoyongLin</dc:creator>
      <dc:date>2023-05-25T21:18:06Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076523#M6133</link>
      <description>&lt;P&gt;What is "&lt;SPAN&gt;&amp;nbsp;"task mode:Incremental synchronization"&amp;nbsp; ?? I installed the May 2023 release just to make sure I did not miss anything new. That release does not add a new taks mode, it does remove the 'cloud' task mode.&lt;/SPAN&gt;&lt;/P&gt;
&lt;P&gt;&lt;SPAN&gt;&amp;gt;&amp;gt; &lt;EM&gt;Why does replicate scan the full table data？&lt;/EM&gt;&lt;/SPAN&gt;&lt;/P&gt;
&lt;P&gt;It does NOT at that stage. That logging tekst is somewhat misleading. Replicate is executing the 'describe_only' option. It actually executes that multiple times (hmmm... not optimal &lt;span class="lia-unicode-emoji" title=":slightly_smiling_face:"&gt;🙂&lt;/span&gt; often in a a single second.&amp;nbsp; See below.&lt;/P&gt;
&lt;P&gt;Please check the timestamps around that statement it should take any time. Next check where the real 'gap'&amp;nbsp; / delay is happening&lt;/P&gt;
&lt;P&gt;Please consider attaching a (&amp;nbsp; portion(s)&amp;nbsp; of the&amp;nbsp; )&amp;nbsp; log for the reptask_xxx.log&amp;nbsp; as .txt with a couple of lines around one of the 'select *' lines and sectons where you think it spends a lot of time.&lt;/P&gt;
&lt;P&gt;Hein.&lt;/P&gt;
&lt;P&gt;&lt;SPAN&gt;&lt;A href="https://docs.oracle.com/cd/A97630_01/appdev.920/a96584/oci04sql.htm" target="_blank"&gt;https://docs.oracle.com/cd/A97630_01/appdev.920/a96584/oci04sql.htm&lt;/A&gt;&lt;/SPAN&gt;&lt;/P&gt;
&lt;P&gt;&lt;SPAN&gt;OCI_DESCRIBE_ONLY. This mode is for users who wish to describe a query prior to execution. Calling OCIStmtExecute() in this mode does not execute the statement, but it does return the select-list description.&lt;/SPAN&gt;&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 21:30:00 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076523#M6133</guid>
      <dc:creator>Heinvandenheuvel</dc:creator>
      <dc:date>2023-05-25T21:30:00Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076529#M6136</link>
      <description>&lt;P&gt;This task has not written data since 03:53:10:824043，the sql statement "SELECT * FROM "SJZBQ_SS"."DZDZ_HWXX_DZFP""&amp;nbsp; has been running on the target database。&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 21:57:08 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076529#M6136</guid>
      <dc:creator>XiaoyongLin</dc:creator>
      <dc:date>2023-05-25T21:57:08Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076533#M6138</link>
      <description>&lt;P&gt;Hi&amp;nbsp;&lt;a href="https://community.qlik.com/t5/user/viewprofilepage/user-id/151740"&gt;@XiaoyongLin&lt;/a&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Please respond to the initial questions on your support case and attach a diagnostics package for the task. If it is the metadata check that is taking time, I do not know of a way to avoid that or speed it up. It might help to ensure the indexes are not fragmented and that the statistics are up to date. Please check with the source/target DBA for those items.&lt;/P&gt;
&lt;P&gt;Thanks,&lt;/P&gt;
&lt;P&gt;Dana&lt;/P&gt;</description>
      <pubDate>Thu, 25 May 2023 22:21:27 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076533#M6138</guid>
      <dc:creator>Dana_Baldwin</dc:creator>
      <dc:date>2023-05-25T22:21:27Z</dc:date>
    </item>
    <item>
      <title>Re: The task starts slowly</title>
      <link>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076589#M6142</link>
      <description>&lt;P&gt;&amp;gt;&amp;gt;&amp;nbsp;&lt;EM&gt;This task has not written data since 03:53:10:824043，the sql statement "SELECT * FROM "SJZBQ_SS"."DZDZ_HWXX_DZFP"" has been running on the &lt;STRONG&gt;target&lt;/STRONG&gt; database。&lt;/EM&gt;&lt;/P&gt;
&lt;P&gt;Did it ever write data? Did you see any "Unload finished for segment ..." with more than 0&amp;nbsp; "rows sent"&lt;/P&gt;
&lt;P&gt;I don't think that query runs against the &lt;STRONG&gt;target&lt;/STRONG&gt; (when doing drop + create). What makes you think is does? Show us!&lt;/P&gt;
&lt;P&gt;Can you see SQL activity on source (or target) corresponding to the Replicate PID/Thread?&lt;/P&gt;
&lt;P&gt;Please show a few more lines around/before&amp;nbsp;&lt;EM&gt;03:53:10:824043 &lt;/EM&gt;as requested.&lt;/P&gt;
&lt;P&gt;I believe is was 'slow', but did it ever finish? Did it ever work at all? Which Replicate version Which Oracle versions?&lt;/P&gt;
&lt;P&gt;Thanks,&lt;/P&gt;
&lt;P&gt;Hein.&lt;/P&gt;</description>
      <pubDate>Fri, 26 May 2023 04:44:30 GMT</pubDate>
      <guid>https://community.qlik.com/t5/Qlik-Replicate/The-task-starts-slowly/m-p/2076589#M6142</guid>
      <dc:creator>Heinvandenheuvel</dc:creator>
      <dc:date>2023-05-26T04:44:30Z</dc:date>
    </item>
  </channel>
</rss>

