 |
 |
Could we improve error messages for the case when scisql is not installed? I'd dare to say it'd even be ok to make some parts die hard, the system is not very usable without scisql (and if it happens it will be during system warmup) Jacek On 09/05/2013 12:36 AM, Bill Chickering wrote: > You're correct, scisql is not installed. At the same time, the chunkQueries are not reaching the workers (no chunk query execution error--doesn't get that far), so there is also an xrootd related issue. > > I see that scisql has dependencies MySQLdb and Mako. Perhaps I should install all three of these within my qserv sandbox. Is there an obvious reason why these aren't included as part of the standard qserv build? > > Thanks, > Bill > > On Sep 4, 2013, at 9:34 PM, Serge Monkewitz <[log in to unmask]> wrote: > >> Hi Bill, >> >> In general, if you get an error like "read/write failed for chunks foo, bar, baz", you should check the worker xrootd log to see the underlying chunk query execution error. >> >> Totally guessing here, but it looks to me like all the failed queries contain scisql_ UDF calls, and all the (semi-) successful ones don't. Is it possible you didn't install scisql for the worker mysql instances? >> >> Cheers, >> Serge > > ######################################################################## > 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 > ######################################################################## 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