Re: Artemis failed but HA secondary did not take over

I can't really say if your proposed configuration is "wise" or not since I don't know all the details of your use-case. However, I can say a few things conclusively... First, the global-max-size is measured in bytes so using 4096 will only give you 4KB of space. If you actually want 4 gigabytes th

Re: Artemis failed but HA secondary did not take over

I see. With 16G on the machine, is this a wise setup? -Xms1024M -Xmx8G 4096 On Wed, Jan 20, 2021 at 1:28 PM Justin Bertram wrote: > > I don't have max-size-bytes or global-max-size set. However, according to > the comments in the broker.xml file, global-max-size defaults to half of > -Xmx, so

Re: Client connection setting to use a particular nic interface and ip address

When you configure the transportConnector on the ActiveMQ Server, set the url to use the IP you want, not the listen-on-all-IPs placeholder “0.0.0.0" > On Jan 20, 2021, at 9:03 AM, Bijju Patel wrote: > > Hi, > The client system from where I am connecting to the broker, has 2 NIC cards > and hen

Re: MQTT in, JMS out

MQTT only supports a bytes message, so ActiveMQ is doing what it can to deliver as an ActiveMQBytesMessage. Once you receive it, it should be a simple new String(bytes) to have the JSON in a string. NOTE— check your character encoding handling to avoid those pain points. UTF-8, etc. > On Jan

Re: Artemis failed but HA secondary did not take over

> I don't have max-size-bytes or global-max-size set. However, according to the comments in the broker.xml file, global-max-size defaults to half of -Xmx, so shouldn't Artemis start using the page file after it uses 1G? Yes. However, given a few recent issues like ARTEMIS-3021 [1] and ARTEMIS-3061

Re: Artemis failed but HA secondary did not take over

Hi Justin, There is 16GB of memory on the VM. And I have JVM memory set to -Xms512M -Xmx2G in the artemis.profile file, so I can increase those. I don't have max-size-bytes or global-max-size set. However, according to the comments in the broker.xml file, global-max-size defaults to half of -Xmx,

RE: Client connection setting to use a particular nic interface and ip address

Hi Bijju. You could set up a static route for the target server. To be able to specify more exactly, I'd need to know more about the client. What platform, and what your IP configuration and route tables look like. Peace... Sridhar > -Original Message- > From: Bijju Patel > Sent:

Re: Artemis failed but HA secondary did not take over

> java.lang.OutOfMemoryError: Java heap space My guess is that you were approaching your heap limit which caused the preceding issues. Once you hit an OutOfMemoryError then in my experience all bets are off so to speak and you need to restart the broker. It's not clear why you couldn't connect to

Client connection setting to use a particular nic interface and ip address

Hi, The client system from where I am connecting to the broker, has 2 NIC cards and hence configured with 2 ip addresses. When I make a connection to broker, I want to make the connection using a particular IP address (and hence using a particular nic) ie the client system has 10.1.10.10 and 10.1.1

MQTT in, JMS out

Hi everybody. I'm working on an integration between AWS CloudWatch and IBM Tivoli Netcool/OMNIbus via its Message Bus Probe. Because of various adventures, I have settled on an architecture using a Lambda to post to a topic in Amazon MQ, which is managed ActiveMQ Classic 5.5.14. The Lambda c

Artemis failed but HA secondary did not take over

I had an incident last night. I can share the actual log files with anyone interested in seeing the full picture. Below is a summary. I also have OS resource tracking data. I have a pair of Artemis servers using a shared store -- an NFS mounted folder. My use case is enterprise integration. I use