Hi JB, agree, I guess I am more used to use ConcurrentHashMap than synchronized keyword. Ok, will close PR.
Kind Regards, Miroslav V V pon., 1. okt. 2018 ob 11:08 je oseba Jean-Baptiste Onofré < j...@nanthrax.net> napisala: > Just to be clear, this fix is on Pax JDBC, but the same fix should be > applied to Pax JMS. Synchronized is IMHO better and consistent more than > using a ConcurrentHashMap. > > Regards > JB > > On 01/10/2018 11:05, Jean-Baptiste Onofré wrote: > > By the way, don't you think this commit: > > > > > https://github.com/ops4j/org.ops4j.pax.jdbc/commit/6bfeccea774195316d4d2382b5235a7c4d1501e0 > > > > already fix the issue ? > > > > I think the synchronized on methods are enough, so I think your PR is > > useless with this commit. > > > > Regards > > JB > > > > On 01/10/2018 08:51, Miroslav Beranič wrote: > >> Hi all, > >> > >> I was getting ( sometimes ) ConcurrentModificationException when using > >> PAX JMS ( 1.0.2 ). It turned out to be synchronization fail in > >> ConnectionFactoryConfigManager. I've made a fix by replace HashMap with > >> ConcurrentHashMap ( and removed synchronized methods ). > >> > >> I write here, as I think Karaf is main user base of the PAX JMS library. > >> > >> Any comment is welcome. Fix is located at: > >> > >> https://github.com/ops4j/org.ops4j.pax.jms/pull/15 > >> > >> P.S: I've also updated the dependencies ( in my local branch, not part > >> of the pull request ) - something to have in mind when/if doing local > >> build ( working with Karaf 4.2.2 ). > >> > >> Kind Regards, > >> Miroslav > >> > >> > >> > >> -- > >> Miroslav Beranič > >> MIBESIS > >> +386(0)40/814-843 > >> miroslav.bera...@mibesis.si <mailto:miroslav.bera...@mibesis.si> > >> http://www.mibesis.si > > > > -- > Jean-Baptiste Onofré > jbono...@apache.org > http://blog.nanthrax.net > Talend - http://www.talend.com > -- Miroslav Beranič MIBESIS +386(0)40/814-843 miroslav.bera...@mibesis.si http://www.mibesis.si