Seems odd that a read-only query would grab exclusive, blocking locks.
Is there other outstanding work in those transactions?
On 3/29/18 2:42 AM, palo.liska wrote:
Hi
did someone saw deadlock during concurrent select count(*) from tableName?
09:15:41,959 [ERROR] o.h.e.j.s.SqlExceptionHelper:146 - A lock could not be
obtained due to a deadlock, cycle of locks and waiters is:
Lock : ROW, SIPSAFQUEUE, (1,10)
Waiting XID : {20748, S} , EPS, /* SELECT COUNT(*) FROM SIPSAFQueue r
*/
select count(*) as col_0_0_ from SIPSAFQueue sipsafqueu0_
Granted XID : {20506, X}
Lock : ROW, SIPSAFQUEUE, (1,14)
Waiting XID : {20506, S} , EPS, /* SELECT COUNT(*) FROM SIPSAFQueue r
*/
select count(*) as col_0_0_ from SIPSAFQueue sipsafqueu0_
Granted XID : {20748, X}
. The selected victim is XID : 20748.
--
Sent from:
http://apache-database.10148.n7.nabble.com/Apache-Derby-Users-f95095.html
