I believe you're correct. If PFC is disabled, when the store cursor checks
limits, it checks only the memory percentage of its own MemoryUsage and
compares it to the high water mark. Otherwise if PFC is enabled, it checks
whether the MemoryUsage is full but the isFull method also checks its
You can also use any of the ActiveMQ 5.x clients, as Apollo supports
OpenWire. The support for OpenWire isn't yet in the complete state, but a
majority of the features are there. I'm working on the initial
documentation, and have a project set up here to demo some of the features:
Hiram, and others...
would it be possible to turn on the JIRA+SVN plugin so it correlates SVN
commits with JIRAs in the Source tab like ActiveMQ has it?
Thanks,
Christian
-
http://www.christianposta.com/blog
--
View this message in context:
@rmn190
I've added a new blog post describing benchmarking JMS Layer, specifically
ActiveMQ:
http://www.christianposta.com/blog/?p=268
-
http://www.christianposta.com/blog
--
View this message in context:
I have seen this, as Torsten mentioned, when the subscriber is
connecting/disconnecting/reconnecting all the time
-
http://www.christianposta.com/blog
--
View this message in context:
http://activemq.2283324.n4.nabble.com/Durable-Subscriber-flashes-when-monitored-tp4655578p4655769.html
This has been fixed in latest trunk, Dean. Feel free to verify it!
You can checkout the latest snapshot builds from here:
http://activemq.apache.org/apollo/download.html
-
http://www.christianposta.com/blog
--
View this message in context:
Yes, please show the code you're using.
-
http://www.christianposta.com/blog
--
View this message in context:
http://activemq.2283324.n4.nabble.com/Message-expiring-before-the-lifetime-when-rollback-is-used-tp4655667p4655687.html
Sent from the ActiveMQ - User mailing list archive at
This might point you in the right direction, but not sure:
http://tmielke.blogspot.com/2011/09/activemq-network-bridge-to-masterslave.html
Is this something that can be recreated consistently?
-
http://www.christianposta.com/blog
--
View this message in context:
This should really be a new thread.
take a look at http://activemq.apache.org/message-cursors.html
you have producer flow control set to true, so your producers will be
throttled if there is no space available. using message cursors, you can get
the messages spooled to disk so the slowest
if you attach your config files i can give it a try on my side
-
http://www.christianposta.com/blog
--
View this message in context:
http://activemq.2283324.n4.nabble.com/Store-and-Forward-with-embedded-brokers-stand-alone-broker-tp4655028p4655431.html
Sent from the ActiveMQ - User mailing
I can recreate this as you've described. Not sure why it's happening. I've
opened a JIRA for it and will hack at it when I get a sec.
https://issues.apache.org/jira/browse/APLO-246
-
http://www.christianposta.com/blog
--
View this message in context:
From your logs:
sk92129 wrote
2012-08-22 12:58:21,363 | INFO | Listening for connections at:
ssl://centos-test1.foo.com:61616?needClientAuth=true |
org.apache.activemq.transport.TransportServerThreadSupport | main
You can see from your config:
sk92129 wrote
The network connector in broker 2 has duplex set to true
This will open a connection in both directions, which explains the random
port on broker1.
Can you try having uni-directional network connectors on each broker?
-
http://www.christianposta.com/blog
--
View this message in context:
any chance you took a heap dump?
-
http://www.christianposta.com/blog
--
View this message in context:
http://activemq.2283324.n4.nabble.com/java-lang-OutOfMemoryError-Java-heap-space-tp4655291p4655364.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
There isn't a timeout for JMSXGroupID as far as I can tell.
I wonder if you could listen to advisories for a particular destination and
log last-seen group ids? and then send the -1 group sequence after your
desired timeout?
-
http://www.christianposta.com/blog
--
View this message in
What version (apollo and/or windows) are you using?
I just tried on WinXP and it works fine with Apollo v1.4
-
http://www.christianposta.com/blog
--
View this message in context:
http://activemq.2283324.n4.nabble.com/Apollo-as-Windows-Service-tp4655390p4655403.html
Sent from the ActiveMQ -
Apollo relies on JAAS LoginModule to handle authentication.
Out of the box, there is support for file-based authentication:
http://svn.apache.org/viewvc/activemq/activemq-apollo/trunk/apollo-broker/src/main/scala/org/apache/activemq/apollo/broker/security/FileUserLoginModule.scala?view=markup
Yes, you can.
Specify an access_rule as mentioned here:
http://activemq.apache.org/apollo/documentation/user-manual.html#Authorization
Use the id attribute to specify your destination, and the kind attribute
to specify the type of destination.
See the Send and create authorized via id_regex
It's targeted for official release later this month or early next (Sept).
-
http://www.christianposta.com
--
View this message in context:
http://activemq.2283324.n4.nabble.com/Fuse-Message-Broker-Based-On-ActiveMQ-5-6-tp4654698p4654706.html
Sent from the ActiveMQ - User mailing list
From what I understand, Apollo uses the idea of consumer credit window to
decide how many messages to send down to the client that haven't been ACKed.
use the credit header in your stomp frame to set the credit window:
I figured it out. The short answer is no. A longer description of how the
network connectors work can be found on my blog where I dumped my notes on
the whole thing: http://www.christianposta.com/blog/?p=242
-
http://www.christianposta.com
--
View this message in context:
quick question about the network connectors/bridge and failover. seems the
static discover is sufficient to handle failovers as it will poll the slave
in a master/slave network of brokers. seems the polling is configurable
also. is there any reason to use the failover() mechanism when establishing
If you give me some more info, I can try and reproduce on my end. Let me know
your config file and how your producers/consumers are connecting.
-
http://www.christianposta.com
--
View this message in context:
Abimael,
Just to follow up on what Matt said, the messages in the log files will only
be deleted once they're *all* consumed. If you have a log file where all but
one message is consumed, the log file will stay around until the last
message is consumed.
Thanks,
Christian
-
Amir,
I figured it out. The patch you submitted was not the correct one, but very
much on the right track. I'll update the JIRA with what I found. As Gary
mentioned, this module could use some unit tests, so I'll write those as
well.
The link to your JIRA issue is here:
Amir,
I see you have a JIRA still opened for this issue. Can you confirm this is
still an issue?
--
View this message in context:
http://activemq.2283324.n4.nabble.com/UDP-Transport-connector-listens-on-a-random-port-number-tp3570237p4241119.html
Sent from the ActiveMQ - User mailing list
Hi all, my first post to this mailing list.
I'd like to get some ideas about what message broker should I use for a
messaging integration system. I realize ActiveMQ may be good for this, but
I'd like to get some feedback from others that may have used Weblogic's JMS
provider (or both) and give me
27 matches
Mail list logo