[ https://issues.jboss.org/browse/TEIID-3404?page=com.atlassian.jira.plugin... ] Ramesh Reddy commented on TEIID-3404: ------------------------------------- I did not wanted to use JDBC as there are enough database specific issues, we need to provide plugins for each source. JPA seemed lot simpler and even can create the tables on the fly. Is batching possible in JPA? if yes, may be we can use it. This mechanism using EJB with JPA, we need to see if this a EJB pooling issue or what is contributing to the high CPU. The thing I do not like about it is there is no further management we offer on it, like archiving, log rotations or viewing or searching, nor I would like to go into that space, that is where below solutions come in.. What would be interesting is move on to new log management systems like http://logstash.net/ http://www.fluentd.org/ https://cwiki.apache.org/confluence/display/FLUME/Home (written in java) http://kafka.apache.org/ I am sure for most of them they may already support the Log4J appenders, we should research into these as separate issue. -- This message was sent by Atlassian JIRA (v6.3.11#6341)