Since this affects all daemons how about a pre-fetch of 1? On Thu, Sep 28, 2017 at 6:48 PM, Jaroslaw Cwiklik <[email protected]> wrote:
> Actually, the correct broker decoration should be: > > # # The broker *client* url decoration (ie - DUCC daemons). > ducc.broker.url.decoration=jms.useCompression=true&jms. > prefetchPolicy.all=0 > > Note: jms.prefetchPolicy.all=0 > > This sets limits for all consumer types (topic and queue consumers) > > The earlier proposed decoration jms.prefetchPolicy.queuePrefetch=0 only > effects > consumers on a queue. Ducc daemons communicate mostly via topics. > > > On Thu, Sep 28, 2017 at 6:36 PM, Jerry Cwiklik (JIRA) <[email protected] > > > wrote: > > > > > [ https://issues.apache.org/jira/browse/UIMA-5587?page= > > com.atlassian.jira.plugin.system.issuetabpanels:comment- > > tabpanel&focusedCommentId=16185023#comment-16185023 ] > > > > Jerry Cwiklik commented on UIMA-5587: > > ------------------------------------- > > > > The fix for the prefetch needs to go into default.ducc.properties > > > > # # The broker *client* url decoration (ie - DUCC daemons). > > ducc.broker.url.decoration=jms.useCompression=true&jms. > > prefetchPolicy.queuePrefetch=0 > > > > Note: jms.prefetchPolicy.queuePrefetch=0 > > > > With the above, a JMS consumer does not cache any msgs and instead asks a > > broker for a single message at a time. This change effects all JMS > > consumers (Ducc daemons). Real services should use application broker > not > > the Ducc broker. I have not delivered any changes yet giving all a chance > > to review this proposal. > > > > > DUCC Process Manager (PM) takes OOM on Apache DUCC demo > > > ------------------------------------------------------- > > > > > > Key: UIMA-5587 > > > URL: https://issues.apache.org/jira/browse/UIMA-5587 > > > Project: UIMA > > > Issue Type: Bug > > > Components: DUCC > > > Reporter: Lou DeGenaro > > > Assignee: Jerry Cwiklik > > > Priority: Minor > > > Fix For: future-DUCC > > > > > > > > > Apache DUCC demo has PM that runs with reduced memory -Xmx200M > > > On at least 2 occasions PM has taken OOM. Core dumps have resulted > that > > indicate that PM is flooded with multiple OR publications that is > supposed > > to be handled by "BlastGuard"? > > > Investigate what storage is being used for and if multiple OR messages > > are being processed in parallel. Perhaps Camel should be configured to > > process OR message sequentially if it isn't already. > > > > > > > > -- > > This message was sent by Atlassian JIRA > > (v6.4.14#64029) > > >
