Hi Daniela

I am a little bit confused now, all this information is in (or linked from) the GGUS ticket.  There was the original ticket from the dashboard[1] and then you asked me to create a separate ticket against the Monitoring group to ask about the test and to link it back to the original, which I did[2].  The discussion with the developers is in that 2nd ticket and you can see the agreed plan of action.  I have even updated the GOCDB with the new setting so it can be picked up as soon as the new test is ready[3].

Alastair

[1] https://ggus.eu/index.php?mode=ticket_info&ticket_id=124876
[2] https://www.ggus.org/index.php?mode=ticket_info&ticket_id=125026 
[3] https://goc.egi.eu/portal/index.php?Page_Type=Service&id=8854





On 24 Jan 2017, at 12:28, Daniela Bauer <[log in to unmask]> wrote:

Hi Alistair,

If you need the test changed, have you told the test developers ?

regards,
Daniela

On 24 January 2017 at 11:47, Alastair Dewhurst <[log in to unmask]> wrote:
Hi Daniela

I didn’t think it was particularly onerous to read the ticket.

In order to pass the test we need to pass an additional option to the probe (--path).  This is not possible at this moment and we are waiting on the developers for a fix.  

I have also spoken to both the VOs Echo will be supporting initially (ATLAS and CMS) and neither of them cared about us passing this test.  They do need Echo registered in the GOCDB so they can pick up scheduled downtimes.  They completely supported my position that the tests should evolve to represent the VO use case.  Echo is going to provide the functionality that the VOs we support need going forward.  For example, we are not providing an SRM and we have replaced the information provider with a 3 line script that uploads a .json file to the storage.  Storage is evolving, some other services will need to do this too and I have been providing them with the input they need to do this.

Alastair


On 24 Jan 2017, at 08:35, Daniela Bauer <00000576bab2846e-dmarc-[log in to unmask]> wrote:

Hi All,

I won't be able to attend the ops meeting today as I am attending a course[*]. At the end of this email  are the notes from the CMS computing meeting.

I'm also on ROD duty this week. I realize this has come up before, but can't RAL (as a T1 no less) not just enable ops tests wrt this ticket and be done with it ?
 https://ggus.eu/index.php?mode=ticket_info&ticket_id=124876
The requirement that machines in production should pass basic tests is really not that
onerous.
There is also an odd failure for an ARC test for a Brunel CE (21), which looks to me more like a castor failure:
https://argo-mon2.egi.eu/nagios/cgi-bin/extinfo.cgi?type=2&host=dc2-grid-21.brunel.ac.uk&service=org.nordugrid.ARC-CE-SRM-result-ops
Before I assign this ticket to Brunel, could someone at the T1 have a look, please ?

Regards,
Daniela


[*] Machine learning by Yandex. In my mind SkyNet now speaks with a Russian accent,,,,

---------- Forwarded message ----------
From: Ken Bloom <[log in to unmask]>
Date: 23 January 2017 at 20:08
Subject: notes from the facilities services/computing operations meetings, 23/1/17
To: [log in to unmask]



*** Discussion title: Tier-2 Activities

Sites should be aware that the LCG tools, such as lcg-cp, are being phased out.  Make sure that you don’t depend on them!  CMS recommends that sites switch to the gfal tools instead, see <https://dmc.web.cern.ch/projects/gfal2-utils&gt;.  No need to with immediately, but on the timescale of later this year (at least for OSG sites), the LCG utilities won’t be available.

Also, sites will need to make progress on IPv6 this year, meaning dual-stacked systems for central services and storage at the sites.  There are no site instructions yet, CMS will have to work on this, starting with a dedicated meeting when practical.  However, there is experience on OSG, the UK, several T1s, and there is a HEPiX working group on the topic.  Sites who have not yet worked on this can volunteer to be guinea pigs as means of testing documentation.  The hope is that this can take place over the next few months.

Please update to the new SpaceMon client; it’s in CVMFS, and see <https://twiki.cern.ch/twiki/bin/view/CMSPublic/SpaceMonSiteAdmin#Step_4_Installation_and_configur&gt; for updated instructions.

This week T2_CN_Beijing got out of the waiting room and T2_IN_TIFR went in.  Two other sites in the waiting room are T2_GR_Ioannina and T2_RU_IHEP.

Here is the count of open tickets at sites:

T2_EE_Estonia   1
T2_FI_HIP       2
T2_IN_TIFR      1
T2_IT_Bari      2
T2_PK_NCP       3
T2_PL_Warsaw    2
T2_RU_INR       1
T2_RU_ITEP      1
T2_RU_PNPI      2
T2_TW_NCHC      1
T2_UK_SGrid_RALPP       1
T2_US_MIT       2
T2_US_Nebraska  1
T2_US_UCSD      2
T2_US_Vanderbilt        1

That’s not too many!  Best wishes.

Ken

-------------------------------------------------------------
Visit this CMS message (to reply or unsubscribe) at:
https://hypernews.cern.ch/HyperNews/CMS/get/t2/1359.html



--
Sent from the pit of despair

-----------------------------------------------------------
[log in to unmask]
HEP Group/Physics Dep
Imperial College
London, SW7 2BW
Tel: +44-(0)20-75947810
http://www.hep.ph.ic.ac.uk/~dbauer/




--
Sent from the pit of despair

-----------------------------------------------------------
[log in to unmask]
HEP Group/Physics Dep
Imperial College
London, SW7 2BW
Tel: +44-(0)20-75947810
http://www.hep.ph.ic.ac.uk/~dbauer/