Skip to main content

Java Garbage Collection

This week we got a plenty of out of heap memory exceptions in Java. So I started looking on Java Garbage Collection, a revision of System Software Internals theory again.  Java has its heap divided into new gen and old generation. Java's garbage collection tries to take advantage of the fact that new object will be deleted sooner(objects will have smaller life time).
So Java's heap is divided into
1)Young Generation(smaller in size)
2)Old Generation (larger chunk)

Young Generation:
Young Generation is smaller in size. Traditional Recursive Garbage collection will be faster if the size to be collected is smaller. New Objects are created in Eden part of the Heap. Garbage Collector runs frequently in this space and marks objects that are referenced.  Objects that are not not referenced are removed in the second pass and the live objects are moved to Survivor Space.
Survivor Space has two regions Survivor 0 and Survivor 1 which uses an algorithm similar to Copying Garbage Collector. So Garbage Collector runs on Survivor(0/1) depending on live region and moves the live objects to the other region and the former region will have only death objects that will be freed. Objects that survive more than two/three rounds of garbage collection will be moved to Old Generation.
Old Generation:
Old Generation has objects which are alive for longer time. Since Java's garbage collection is based on the fact that objects tend to die sooner. Old Generation is collected very rarely and it is very huge in size. The old generation runs two algorithms
1)CMS-Concurrent Mark Sweep
This marks all objects that are not alive and flushes them but there will be defragmentation across the heap
2)Stop the World Collector
Stop the world collector kicks in when heap is almost full, it compacts the live objects. This will cause the java process to hang for few seconds and cause performance impact especially when dealing with heap space as high as 20/30 GB as in our case. 
Java 7 has come up with G1 collector which optimizes by combining CMS and STW collector. 

To see the garbage collection live, there is a good tool jstat
jstat -gc <pid> <interval to monitor>


Popular posts from this blog

How we have systematically improved the roads our packets travel to help data imports and exports flourish

This blog post is an account of how we have toiled over the years to improve the throughput of our interDC tunnels. I joined this company around 2012. We were scaling aggressively then. We quickly expanded to 4 DCs with a mixture of AWS and colocation. Our primary DC is connected to all these new DCs via IPSEC tunnels established from SRX. The SRX model we had, had an IPSEC throughput of 350Mbps. Around December 2015 we saturated the SRX. Buying SRX was an option on the table. Buying one with 2Gbps throughput would have cut the story short. The tech team didn't see it happening. I don't have an answer to the question, "Is it worth spending time in solving a problem if a solution is already available out of box?" This project helped us in improving our critical thinking and in experiencing the theoretical network fundamentals on live traffic, but also caused us quite a bit of fatigue due to management overhead. Cutting short the philosophy, lets jump to the story.

LXC and Host Crashes

 We had set up a bunch of lxc containers on two servers each with 16 core CPUs and 64 GB RAM(for reliability and loadbalancing). Both the servers are on same vlan. The servers need to have atleast one of their network interface in promiscuous mode so that it forwards all packets on vlan to the bridge( ) which takes care of the routing to containers. If the packets are not addressed to the containers, the bridge drops the packet. Having this setup, we moved all our platform maintenance services to these containers. They are fault tolerant as we used two host machines where each host machine has a replica of the containers on the other. The probability to crash for both the servers at the same time due to some hardware/software failure is less. But to my surprise both the servers are crashing exactly the same time with a mean life time 20 days. We had to wake up late nights(early mornings) to fix stuffs that gone down The

The server, me and the conversation

We were moving a project from AWS to our co-located DC. We have setup KVMs scheduled by Cloudstack for each of the component in the architecture. The KVMs used local storage. The VMs are provisioned with more than required resources because we have the opinion that in our DC scaling during peak load and then downscaling doesn't offer much benefits financially as we are anyways paying for the hardware in advance and its also powered on. Its going to be idle if not used. Now we found something interesting our latency in co-located DC was 2 times more than in AWS. The time for first byte at our load balancer in aws was 60ms average and at our DC was 112ms. We started our debugging mission, Mission Conquer-AWS. All the servers are newer Dell hardwares. So the initially intuition was virtualisation is causing the issue. Conversation with the Hypervisor We started with CPU optimisation, we started using the host-passthrough mode of CPU in libvirt so VMs dont see QEMU emulated CPUs,