Документ взят из кэша поисковой машины. Адрес оригинального документа : http://theory.sinp.msu.ru/pipermail/ru-ngi/attachments/20140623/5f05831e/attachment.html
Дата изменения: Mon Jun 23 12:47:29 2014
Дата индексирования: Mon Apr 11 01:41:18 2016
Кодировка: koi8-r
<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=KOI8-R">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Т2 admins,<br>
    Спасибо всем за инфо.<br>
    Жаль, что мы так и не услышали начальника T2_RU_RRCKI.<br>
    Вот что я послал Оливеру Гутче в овет на его просьбу: <br>
    <div class="moz-forward-container"><br>
      <br>
      -------- Original Message --------
      <table class="moz-email-headers-table" border="0" cellpadding="0"
        cellspacing="0">
        <tbody>
          <tr>
            <th align="RIGHT" nowrap="nowrap" valign="BASELINE">Subject:
            </th>
            <td>T2_RU_* low readiness rank</td>
          </tr>
          <tr>
            <th align="RIGHT" nowrap="nowrap" valign="BASELINE">Date: </th>
            <td>Mon, 23 Jun 2014 12:39:21 +0400</td>
          </tr>
          <tr>
            <th align="RIGHT" nowrap="nowrap" valign="BASELINE">From: </th>
            <td>Victor Zhiltsov <a class="moz-txt-link-rfc2396E" href="mailto:jiltsov@cern.ch">&lt;jiltsov@cern.ch&gt;</a></td>
          </tr>
          <tr>
            <th align="RIGHT" nowrap="nowrap" valign="BASELINE">Reply-To:
            </th>
            <td><a class="moz-txt-link-abbreviated" href="mailto:jiltsov@cern.ch">jiltsov@cern.ch</a></td>
          </tr>
          <tr>
            <th align="RIGHT" nowrap="nowrap" valign="BASELINE">Organization:
            </th>
            <td>CERN</td>
          </tr>
          <tr>
            <th align="RIGHT" nowrap="nowrap" valign="BASELINE">To: </th>
            <td><a class="moz-txt-link-abbreviated" href="mailto:hn-cms-comp-ops@cern.ch">hn-cms-comp-ops@cern.ch</a></td>
          </tr>
        </tbody>
      </table>
      <br>
      <br>
      <pre>Oli,
Here is the info from T2_RU_* admins regarding their low readiness rank:

==============================================================
T2_RU_IHEP
Seems to be OK

T2_RU_JINR
OK

T2_RU_SINP
We are out of "Production" state for uncertain time because of air 
conditioning system malfunction.

T2_RU_RRCKI
No answer.

T2_RU_PNPI
At the end of March two tickets were opened for T2_RU_PNPI:
   <a class="moz-txt-link-freetext" href="https://savannah.cern.ch/support/?142774">https://savannah.cern.ch/support/?142774</a>
   <a class="moz-txt-link-freetext" href="https://savannah.cern.ch/support/?142852">https://savannah.cern.ch/support/?142852</a>
But CMS support does not answer any questions regarding the info on 
crashing jobs.

T2_RU_ITEP
Last time we often come across the error of type
   CRITICAL: [Waiting-&gt;Cancelled [timeout/dropped]] 'BrokerHelper: no
   compatible resources'. 
<a class="moz-txt-link-freetext" href="https://wms305.cern.ch:9000/xdKhZitly5AOFO_DQka10A">https://wms305.cern.ch:9000/xdKhZitly5AOFO_DQka10A</a>
To all tickets and questions we reply that this is not our fault, but 
CMS BDII, and supply the results of our tests.
We see no continuation of the dialog.

T2_RU_INR
After upgrade of DPM to version 1.8.8 и dpm-dsi and dpm-dsi-1.9.3-3.
<a class="moz-txt-link-freetext" href="https://cmsweb.cern.ch/phedex/prod/Activity::QualityPlots?graph=quality_all&amp;entity=link&amp;src_filter=&amp;dest_filter=T2_RU_INR&amp;no_mss=true&amp;period=l48h&amp;upto=&amp;.submit=Update">https://cmsweb.cern.ch/phedex/prod/Activity::QualityPlots?graph=quality_all&amp;entity=link&amp;src_filter=&amp;dest_filter=T2_RU_INR&amp;no_mss=true&amp;period=l48h&amp;upto=&amp;.submit=Update</a>
all turned red.
After changing all ConfigPart.FTSDownload* at all places on
 -service         <a class="moz-txt-link-freetext" href="https://lcgfts3.gridpp.rl.ac.uk:8443">https://lcgfts3.gridpp.rl.ac.uk:8443</a>
only some items turned green.
==============================================================

As it turns out, if we exclude the HW problem at SINP, the problem in 
general is caused primarily by the lack of communication 
between T2 sites and CMS.

That is - the performance of T2_RU_* sites in general may actually be 
not as poor as it can be seen in readiness plots.

-- 
--
Victor Zhiltsov
CERN 22 1-019 x78846
Mob: +41-79-527-1960

</pre>
      <br>
    </div>
    <br>
  </body>
</html>