 |
 |
On 12/17/14 23:13 , Jacek Becla wrote: > I've been mentioning high level spreadsheet with > long term plan. It is in docushare, LDM-240, see: > > https://docushare.lsstcorp.org/docushare/dsweb/Get/LDM-240/LDM-240%2020141118-outlined.xlsx > Btw, I know Jacek knows this, but others may not -- for any LDM-* document (or L??-* in general), you can use: http://ls.st/LDM-NNN (to get the document) http://ls.st/LDM-NNN* (to list all previous and draft versions) (also, the magic URLs above are case insensitive). - M. > > rows 40-48 are in particular relevant for us. > > From qserv perspective, this is what we are on a hook to do in FY 15 > > ----- > Rewrite build and installation scripts and test harness > Evaluate alternative systems and components > Begin rewrite and restructuring of master, worker, and metadata server > Rewrite master and worker and metadata server to support: > Creation and removal of distributed databases > Creation and loading of prepartitioned distributed tables > Initial version of query aborting > Implement initial query types > All of the above robust to user errors, crash-free, and fault-tolerant > ---- > > Also, have a look at KPIs, we promised to demonstrate: > > LV query rate simult. queries qserv 50 > LV query response time seconds qserv 18 > HV shared scan duration hours qserv 24 > HV simultaneous queries queries qserv 6 > > We HAVE TO demonstrate the KPIs by 9/30/2015 > > Jacek > > ######################################################################## > Use REPLY-ALL to reply to list > > To unsubscribe from the QSERV-L list, click the following link: > https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=QSERV-L&A=1 -- Mario Juric Web : http://research.majuric.org Phone : +1 609 933 1033 ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the QSERV-L list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=QSERV-L&A=1