<?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: Performance problem with syslog in multithreaded application in Operating System - HP-UX</title>
    <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545000#M839331</link>
    <description>Thanks for your suggestion. I can see a slight improvement after calling pthread_mutex_setyieldfreq_np(PTHREAD_MUTEX_YIELDNEVER_NP).&lt;BR /&gt;&lt;BR /&gt;But it is not even close to the response times without the syslog calls. &lt;BR /&gt;What is syslog doing here? Maybe I should use another syslog client implementation?</description>
    <pubDate>Mon, 16 May 2005 07:51:37 GMT</pubDate>
    <dc:creator>Nils-Olof Wilske</dc:creator>
    <dc:date>2005-05-16T07:51:37Z</dc:date>
    <item>
      <title>Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544996#M839327</link>
      <description>Hi,&lt;BR /&gt;&lt;BR /&gt;I recently discovered that syslog calls in a multithreaded application is a real performance killer. I guess there is a mutex in the syslog function that blocks the threads, but there must be something more than that.&lt;BR /&gt;If I replace the syslog function with a function that does a  "mutex.lock; printf; mutex.unlock" sequence there is no noticable performance loss.&lt;BR /&gt;I get the same results on HP-UX 11.00 and HP-UX 11.23 (Itanium).&lt;BR /&gt;&lt;BR /&gt;Any ideas?</description>
      <pubDate>Mon, 16 May 2005 03:28:17 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544996#M839327</guid>
      <dc:creator>Nils-Olof Wilske</dc:creator>
      <dc:date>2005-05-16T03:28:17Z</dc:date>
    </item>
    <item>
      <title>Re: Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544997#M839328</link>
      <description>Hi,&lt;BR /&gt;&lt;BR /&gt;It may be interesting to get a 'tusc' of the application with  the options&lt;BR /&gt;-E -v -l -T ""  -f -p -o result.out&lt;BR /&gt;It can give an idea where the time is passed.&lt;BR /&gt;</description>
      <pubDate>Mon, 16 May 2005 03:38:39 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544997#M839328</guid>
      <dc:creator>Laurent Menase</dc:creator>
      <dc:date>2005-05-16T03:38:39Z</dc:date>
    </item>
    <item>
      <title>Re: Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544998#M839329</link>
      <description>As requested I have attached the output of&lt;BR /&gt;'tusc -E -v -l -T "" -f -p'&lt;BR /&gt;for the process. (gzipped)&lt;BR /&gt;A short explanation:&lt;BR /&gt;The process serves clients (8 in this case) connecting via TCP/IP. A thread is created for each client.&lt;BR /&gt;The thread sends a query to Oracle, and returns the result to the client. The thread also makes 6 rapid syslog calls.&lt;BR /&gt;The response time for a client is proportional to the number of clients! Without the syslog calls it is more or less independent of the number of clients.&lt;BR /&gt;</description>
      <pubDate>Mon, 16 May 2005 04:18:10 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544998#M839329</guid>
      <dc:creator>Nils-Olof Wilske</dc:creator>
      <dc:date>2005-05-16T04:18:10Z</dc:date>
    </item>
    <item>
      <title>Re: Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544999#M839330</link>
      <description>Apparantly it sched_yield a lot to get the mutex.&lt;BR /&gt;I suppose it is the cause of the perf kill&lt;BR /&gt;&lt;BR /&gt;you can try to call  not portable&lt;BR /&gt;pthread_mutex_setyieldfreq_np( PTHREAD_MUTEX_YIELDNEVER_NP);&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Mon, 16 May 2005 06:00:09 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3544999#M839330</guid>
      <dc:creator>Laurent Menase</dc:creator>
      <dc:date>2005-05-16T06:00:09Z</dc:date>
    </item>
    <item>
      <title>Re: Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545000#M839331</link>
      <description>Thanks for your suggestion. I can see a slight improvement after calling pthread_mutex_setyieldfreq_np(PTHREAD_MUTEX_YIELDNEVER_NP).&lt;BR /&gt;&lt;BR /&gt;But it is not even close to the response times without the syslog calls. &lt;BR /&gt;What is syslog doing here? Maybe I should use another syslog client implementation?</description>
      <pubDate>Mon, 16 May 2005 07:51:37 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545000#M839331</guid>
      <dc:creator>Nils-Olof Wilske</dc:creator>
      <dc:date>2005-05-16T07:51:37Z</dc:date>
    </item>
    <item>
      <title>Re: Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545001#M839332</link>
      <description>Are you calling openlog() before you spawn your threads, or afterwards (i.e. once per thread)?</description>
      <pubDate>Mon, 16 May 2005 11:01:24 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545001#M839332</guid>
      <dc:creator>Stephen Keane</dc:creator>
      <dc:date>2005-05-16T11:01:24Z</dc:date>
    </item>
    <item>
      <title>Re: Performance problem with syslog in multithreaded application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545002#M839333</link>
      <description>&amp;gt;&amp;gt; Are you calling openlog() before you spawn your threads, &lt;BR /&gt;&amp;gt;&amp;gt; or afterwards (i.e. once per thread)?&lt;BR /&gt;&lt;BR /&gt;openlog() is called once in the main thread before starting any other threads.&lt;BR /&gt;</description>
      <pubDate>Tue, 17 May 2005 01:27:41 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/performance-problem-with-syslog-in-multithreaded-application/m-p/3545002#M839333</guid>
      <dc:creator>Nils-Olof Wilske</dc:creator>
      <dc:date>2005-05-17T01:27:41Z</dc:date>
    </item>
  </channel>
</rss>

