<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <title>"Number of nodes/links read unequal to nodes/links in model" in parallel</title>
  <link rel="self" href="https://dlt-acc.firelay.cloud/c/message_boards/find_thread?p_l_id=3228466&amp;threadId=5026698" />
  <subtitle>"Number of nodes/links read unequal to nodes/links in model" in parallel</subtitle>
  <id>https://dlt-acc.firelay.cloud/c/message_boards/find_thread?p_l_id=3228466&amp;threadId=5026698</id>
  <updated>2026-05-12T12:32:47Z</updated>
  <dc:date>2026-05-12T12:32:47Z</dc:date>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027344" />
    <author>
      <name>Victor Chavarrias</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027344</id>
    <updated>2022-04-20T09:18:00Z</updated>
    <published>2022-04-20T09:18:00Z</published>
    <summary type="html">I am not fully sure I am understanding your workflow. The &amp;lt;icgsolver&amp;gt; flag is set when partitioning, as this is the step when the mdu-files are created. Is the partitioning succesfull? Do you see all the partitioned mdu-files with their correct restart file, partition-polygon-file, and &amp;lt;icgsolver&amp;gt; flag?</summary>
    <dc:creator>Victor Chavarrias</dc:creator>
    <dc:date>2022-04-20T09:18:00Z</dc:date>
  </entry>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027142" />
    <author>
      <name>Niraj Bal Tamang</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027142</id>
    <updated>2022-04-20T07:04:45Z</updated>
    <published>2022-04-20T07:04:45Z</published>
    <summary type="html">Oh. May be this is where i might have made the mistake. I was using (run_dflowfm.bat &amp;#34;--partition:ndomains=36:genpolygon=1:icgsolver=6&amp;#34; &amp;lt;mdu file&amp;gt; &amp;lt;partition polygon file&amp;gt;) command for partitioning. So, does that mean i need to remove the icgsolver from the command and run it first to get the partition and again run it after adding the icgsolver part? </summary>
    <dc:creator>Niraj Bal Tamang</dc:creator>
    <dc:date>2022-04-20T07:04:45Z</dc:date>
  </entry>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027138" />
    <author>
      <name>Victor Chavarrias</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027138</id>
    <updated>2022-04-20T06:54:03Z</updated>
    <published>2022-04-20T06:54:03Z</published>
    <summary type="html">I would do the partitioning already in the cluster immediately before running. &amp;lt;icgsolver&amp;gt; needs to be set to 6 (petsc) for parallel in Linux.</summary>
    <dc:creator>Victor Chavarrias</dc:creator>
    <dc:date>2022-04-20T06:54:03Z</dc:date>
  </entry>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027100" />
    <author>
      <name>Niraj Bal Tamang</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027100</id>
    <updated>2022-04-20T06:06:08Z</updated>
    <published>2022-04-20T06:06:08Z</published>
    <summary type="html">Thank you. I will give it a try. One more question regarding the icgsolver while partitioning. If I am partitioning the file in command prompt and later uploading them in Linux system, should the icgsolver be assigned as 6 or 7? </summary>
    <dc:creator>Niraj Bal Tamang</dc:creator>
    <dc:date>2022-04-20T06:06:08Z</dc:date>
  </entry>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027046" />
    <author>
      <name>Victor Chavarrias</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027046</id>
    <updated>2022-04-20T05:37:00Z</updated>
    <published>2022-04-20T05:37:00Z</published>
    <summary type="html">The partitioned hydrodynamic run will create restart-files such as &amp;lt;simname_0000_20181112_000000_rst.nc&amp;gt;. Be sure to create the partition-polygon file (e.g. &amp;lt;simname_part.pol&amp;gt;).&lt;br /&gt;&lt;br /&gt;For restarting in parallel:&lt;ol style="list-style: decimal outside;" start="1"&gt;&lt;li&gt;Create a new folder for the morphodynamic run. &lt;/li&gt;&lt;li&gt;Create the morphodynamic mdu-file in that folder. &lt;/li&gt;&lt;li&gt;Add the partition-polygon file to the folder.&lt;/li&gt;&lt;li&gt;Create a folder inside called, for instance, &amp;lt;rst&amp;gt;. &lt;/li&gt;&lt;/ol&gt;In the mdu-file of the morphodynamic simulation, refer to the restart-files in flag &amp;lt;RestartFile&amp;gt; in block [restart] as:&lt;br /&gt;&lt;br /&gt;&lt;div class="lfr-code"&gt;&lt;table&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td class="line-numbers"&gt;1&lt;/td&gt;&lt;td class="lines"&gt;&lt;div class="line"&gt;[restart]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td class="line-numbers"&gt;2&lt;/td&gt;&lt;td class="lines"&gt;&lt;div class="line"&gt;RestartFile                       = ./rst/simname_20181112_000000_rst.nc &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt;&lt;/div&gt;&lt;br /&gt;&lt;br /&gt;In the mdu-file of the morphodynamic simulation, refer to the partition-polygon file in flag &amp;lt;PartitionFile&amp;gt; in block [geometry] as:&lt;br /&gt;&lt;br /&gt;&lt;div class="lfr-code"&gt;&lt;table&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td class="line-numbers"&gt;1&lt;/td&gt;&lt;td class="lines"&gt;&lt;div class="line"&gt;[geometry]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td class="line-numbers"&gt;2&lt;/td&gt;&lt;td class="lines"&gt;&lt;div class="line"&gt;PartitionFile                     = simname_part.pol &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt;&lt;/div&gt;</summary>
    <dc:creator>Victor Chavarrias</dc:creator>
    <dc:date>2022-04-20T05:37:00Z</dc:date>
  </entry>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027042" />
    <author>
      <name>Niraj Bal Tamang</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027042</id>
    <updated>2022-04-20T05:20:20Z</updated>
    <published>2022-04-20T05:20:20Z</published>
    <summary type="html">Yes. I tried both ways as mentioned in the manual (using partitioned restart files as well as merging them and using a single restart file). I used the same polygon file to partition the morphodynamic model. I ran the hydrodynamic model in both partitioned and single run just to make the parallel run of morphodynamic model work. I tried both ways one at a time in the morphodynamic model but it didn&amp;#39;t work. But the model is working without partitioning although takes a longer time.   </summary>
    <dc:creator>Niraj Bal Tamang</dc:creator>
    <dc:date>2022-04-20T05:20:20Z</dc:date>
  </entry>
  <entry>
    <title>RE: "Number of nodes/links read unequal to nodes/links in model" in paralle</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027037" />
    <author>
      <name>Victor Chavarrias</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5027037</id>
    <updated>2022-04-20T05:12:56Z</updated>
    <published>2022-04-20T05:12:56Z</published>
    <summary type="html">The error seems to indicate that you are restarting a partitioned run. Is this true? The restart-files need to match each partition. Are the restart-files from a hydrodynamic simulation with the same partitioning?</summary>
    <dc:creator>Victor Chavarrias</dc:creator>
    <dc:date>2022-04-20T05:12:56Z</dc:date>
  </entry>
  <entry>
    <title>"Number of nodes/links read unequal to nodes/links in model" in parallel</title>
    <link rel="alternate" href="https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5026697" />
    <author>
      <name>Niraj Bal Tamang</name>
    </author>
    <id>https://dlt-acc.firelay.cloud/c/message_boards/find_message?p_l_id=3228466&amp;messageId=5026697</id>
    <updated>2022-04-19T23:28:06Z</updated>
    <published>2022-04-19T23:28:06Z</published>
    <summary type="html">My morphodynamic model works fine when run in a single file but when I partition, it is giving me different errors every time. the recent one is the   &amp;#34;Number of nodes/links read unequal to nodes/links in model&amp;#34; at the start of the run. I think I am making a mistake from the partition step to running parallel in the supercomputer. Can anyone please share the steps of partitioning morphodynamic model  including input file feed, so that i can check where I am making the mistake?&lt;br /&gt;Thank you </summary>
    <dc:creator>Niraj Bal Tamang</dc:creator>
    <dc:date>2022-04-19T23:28:06Z</dc:date>
  </entry>
</feed>
