{
  "id": 52861,
  "title": "Anyone else having trouble with timeouts?",
  "url": "/competitions/talkingdata-adtracking-fraud-detection/discussion/52861",
  "author_name": "",
  "post_date": "2018-03-24T04:50:41.540003700Z",
  "votes": null,
  "comment_count": 7,
  "views": 0,
  "content": "<p>I created my first kernel but I am unable to run it as it always reaches max. time limit in Kaggle of 5 hours. Anyone else experienced this?</p>",
  "messages": [
    {
      "id": "302460",
      "postDate": "03/24/2018 04:50:41",
      "content": "<p>I created my first kernel but I am unable to run it as it always reaches max. time limit in Kaggle of 5 hours. Anyone else experienced this?</p>",
      "rawMarkdown": "I created my first kernel but I am unable to run it as it always reaches max. time limit in Kaggle of 5 hours. Anyone else experienced this?",
      "votes": null
    },
    {
      "id": "302483",
      "postDate": "03/24/2018 06:14:48",
      "content": "<pre><code>The kernel was stopped for exceeding the limits on idle time (60 minutes)\n</code></pre>\n\n<p>Happening to my kernels since last day, I guess kaggle has reverted back to 60 mins time limit.</p>",
      "rawMarkdown": "The kernel was stopped for exceeding the limits on idle time (60 minutes)\n\nHappening to my kernels since last day, I guess kaggle has reverted back to 60 mins time limit.",
      "votes": null
    },
    {
      "id": "302573",
      "postDate": "03/24/2018 10:49:17",
      "content": "<p>I may be wrong but based on your wording:  <code>always reaches</code> , I assume that the error is not related to time limit but memory. There are two different limits. </p>\n\n<ul>\n<li>maximum <strong>run time</strong> : max 5/6 hours</li>\n<li>maximum <strong>memory limit</strong> : ~17 GB</li>\n</ul>\n\n<p>You can also identify the error by code. <strong>137</strong> is for memory limit. If the kernel has stopped after error, switch to <strong>log</strong>  tab and you will find detailed error message. If you're seeing following message then probably you're having large training data:</p>\n\n<pre><code>The kernel was killed for trying to exceed the memory limit of 17179869184\n</code></pre>",
      "rawMarkdown": "I may be wrong but based on your wording:  `always reaches` , I assume that the error is not related to time limit but memory. There are two different limits. \n\n - maximum **run time** : max 5/6 hours\n - maximum **memory limit** : ~17 GB\n\nYou can also identify the error by code. **137** is for memory limit. If the kernel has stopped after error, switch to **log**  tab and you will find detailed error message. If you're seeing following message then probably you're having large training data:\n\n    The kernel was killed for trying to exceed the memory limit of 17179869184",
      "votes": null
    },
    {
      "id": "302578",
      "postDate": "03/24/2018 11:21:57",
      "content": "<p>Thanks for your reply Pranav. It doesn't say anything about memory limit although error code is same 137. It says kernel was killed for running more than 21600 seconds.</p>",
      "rawMarkdown": "Thanks for your reply Pranav. It doesn't say anything about memory limit although error code is same 137. It says kernel was killed for running more than 21600 seconds.",
      "votes": null
    },
    {
      "id": "302581",
      "postDate": "03/24/2018 11:37:27",
      "content": "<p>My problem here is not related to idle time but run time.</p>",
      "rawMarkdown": "My problem here is not related to idle time but run time.",
      "votes": null
    },
    {
      "id": "302590",
      "postDate": "03/24/2018 11:50:30",
      "content": "<p>I am not having timeout issue anymore.</p>",
      "rawMarkdown": "I am not having timeout issue anymore.",
      "votes": null
    },
    {
      "id": "302697",
      "postDate": "03/24/2018 15:45:32",
      "content": "<p>also a 1 hr limit on idle time, which is what @SohaibOmar encountered. But I think maybe a kernel can time out for idle time and still be running in the background. In general you can lose interactive contact with a running kernel in various ways (including system glitches), but you can still see the results by loading the kernel when it finishes.</p>",
      "rawMarkdown": "also a 1 hr limit on idle time, which is what @SohaibOmar encountered. But I think maybe a kernel can time out for idle time and still be running in the background. In general you can lose interactive contact with a running kernel in various ways (including system glitches), but you can still see the results by loading the kernel when it finishes.",
      "votes": null
    },
    {
      "id": "353205",
      "postDate": "07/06/2018 06:23:35",
      "content": "<p>Got the same error: The kernel was killed for running longer than 21600 seconds.  exit code:137</p>",
      "rawMarkdown": "Got the same error: The kernel was killed for running longer than 21600 seconds.  exit code:137",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 302483,
      "author_name": "sohaibomar",
      "author_url": "",
      "post_date": "03/24/2018 06:14:48",
      "content": "<pre><code>The kernel was stopped for exceeding the limits on idle time (60 minutes)\n</code></pre>\n\n<p>Happening to my kernels since last day, I guess kaggle has reverted back to 60 mins time limit.</p>",
      "votes": null,
      "replies": [
        {
          "id": 302581,
          "author_name": "abhigrover",
          "author_url": "",
          "post_date": "03/24/2018 11:37:27",
          "content": "<p>My problem here is not related to idle time but run time.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 302590,
          "author_name": "sohaibomar",
          "author_url": "",
          "post_date": "03/24/2018 11:50:30",
          "content": "<p>I am not having timeout issue anymore.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 302573,
      "author_name": "pranav84",
      "author_url": "",
      "post_date": "03/24/2018 10:49:17",
      "content": "<p>I may be wrong but based on your wording:  <code>always reaches</code> , I assume that the error is not related to time limit but memory. There are two different limits. </p>\n\n<ul>\n<li>maximum <strong>run time</strong> : max 5/6 hours</li>\n<li>maximum <strong>memory limit</strong> : ~17 GB</li>\n</ul>\n\n<p>You can also identify the error by code. <strong>137</strong> is for memory limit. If the kernel has stopped after error, switch to <strong>log</strong>  tab and you will find detailed error message. If you're seeing following message then probably you're having large training data:</p>\n\n<pre><code>The kernel was killed for trying to exceed the memory limit of 17179869184\n</code></pre>",
      "votes": null,
      "replies": [
        {
          "id": 302578,
          "author_name": "abhigrover",
          "author_url": "",
          "post_date": "03/24/2018 11:21:57",
          "content": "<p>Thanks for your reply Pranav. It doesn't say anything about memory limit although error code is same 137. It says kernel was killed for running more than 21600 seconds.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 302697,
          "author_name": "aharless",
          "author_url": "",
          "post_date": "03/24/2018 15:45:32",
          "content": "<p>also a 1 hr limit on idle time, which is what @SohaibOmar encountered. But I think maybe a kernel can time out for idle time and still be running in the background. In general you can lose interactive contact with a running kernel in various ways (including system glitches), but you can still see the results by loading the kernel when it finishes.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 353205,
          "author_name": "stephenkung",
          "author_url": "",
          "post_date": "07/06/2018 06:23:35",
          "content": "<p>Got the same error: The kernel was killed for running longer than 21600 seconds.  exit code:137</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "302460": "I created my first kernel but I am unable to run it as it always reaches max. time limit in Kaggle of 5 hours. Anyone else experienced this?",
    "302483": "The kernel was stopped for exceeding the limits on idle time (60 minutes)\n\nHappening to my kernels since last day, I guess kaggle has reverted back to 60 mins time limit.",
    "302573": "I may be wrong but based on your wording:  `always reaches` , I assume that the error is not related to time limit but memory. There are two different limits. \n\n - maximum **run time** : max 5/6 hours\n - maximum **memory limit** : ~17 GB\n\nYou can also identify the error by code. **137** is for memory limit. If the kernel has stopped after error, switch to **log**  tab and you will find detailed error message. If you're seeing following message then probably you're having large training data:\n\n    The kernel was killed for trying to exceed the memory limit of 17179869184",
    "302578": "Thanks for your reply Pranav. It doesn't say anything about memory limit although error code is same 137. It says kernel was killed for running more than 21600 seconds.",
    "302581": "My problem here is not related to idle time but run time.",
    "302590": "I am not having timeout issue anymore.",
    "302697": "also a 1 hr limit on idle time, which is what @SohaibOmar encountered. But I think maybe a kernel can time out for idle time and still be running in the background. In general you can lose interactive contact with a running kernel in various ways (including system glitches), but you can still see the results by loading the kernel when it finishes.",
    "353205": "Got the same error: The kernel was killed for running longer than 21600 seconds.  exit code:137"
  },
  "source": "meta"
}