Skip to main content

Java and lsof

Update: I found that using Runtime.exec was a bad idea because if you have a 2GB VM footprint then the forked process would require 2G free memory in order to run the lsof command.  We had earlier writte a simple python http rpc server that would allow us to execute native commands(like creating a hardlink or running gunzip) from Java and I changed this code to delagate to RPC call few days back.
So the new code looks like



 public void writeTopCommandOutput(Writer writer) throws IOException {
  String rpcRes = Util.doCommandRpc(rpcUrl, "", ListUtil.create("top", "-n", "2", "-b", "-d", "0.2"));
  writer.write(rpcRes);
 }

 public void writeLsofOutput(Writer writer) throws IOException {
  String pid = getJvmProcessid();
  if (pid != null) {
   pid = pid.trim();
   String rpcRes = Util.doCommandRpc(rpcUrl, "", ListUtil.create("lsof", "-p", pid));
   writer.write(rpcRes);
  }
 }


 --------------------------------------
I was chasing a customer issue and some threads in threaddump showed that they are stuck doing filer I/O. Now I needed to chase what file paths these threads are accessing as few days ago this customer had expereinced same issue. So I asked operations to give me lsof output and as we are a distribured team and engineers dont have access to production machines.It always takes time to chase people and to a programmer this means lots of context switches and it derails your thought process.  My goal is to eliminate as many hoops in my debugging path so I wrote a JSP to get me lsof output from java. This will be a jsp accessible through internal ips only, hurray from next release onwards one more reason to avoid Operations team in chasing issues.
Here is the method I added in jsp


    public void writeLsofOutput(Writer writer) throws IOException {
        String pid = getJvmProcessid();
        if(pid!=null){
            pid = pid.trim();
            String cmdLsof = "lsof -p " + pid;
            executeCmd(writer, cmdLsof);            
        }
    }

    private void executeCmd(Writer writer, String cmd) throws IOException {
        String output;
        Process child = Runtime.getRuntime().exec(cmd, new String []{});            
        BufferedReader stdInput = new BufferedReader(new InputStreamReader(child.getInputStream()));
        while ((output = stdInput.readLine()) != null) {
            writer.write(output + "\n");
            writer.flush();
        }
        stdInput.close();
    }
    public String getJvmProcessid() throws IOException {
        String pid = null;
        File pidFile = new File(System.getenv("CATALINA_PID"));
        if(pidFile.exists()) {
            FileInputStream fin = new FileInputStream(pidFile);
            List lines = IOUtils.readLines(fin);
            fin.close();
            pid = StringUtils.join(lines.toArray());
        }
        return pid;
    } 

Comments

  1. I am surprised that your code review team is allowing you to put in such a jsp. Usually a fork/exec system command run from inside the jvm is not a good idea. Also, if you have some other problem with your java server, this lsof will not return causing more threads to be consumed in your java server waiting for this to finish.

    I think you should use some other process for generating this output.

    -Chris

    ReplyDelete
  2. Chris,

    Thanks for reviewing, I agree that in some situations like when a mount point is totally toast lsof might not return, in that case anyways there would be other tomcat threads that would be stuck in filer I/O and that tomcat would anyways need a bounce in maintenance window to come to a sane state. This JSP will be called in some rare situations and in most of the situations lsof should return and in case the first execution of jsp is stuck, I wont call another one to avoid tomcat threads being gobbled up.

    ReplyDelete

Post a Comment

Popular posts from this blog

Haproxy and tomcat JSESSIONID

One of the biggest problems I have been trying to solve at our startup is to put our tomcat nodes in HA mode. Right now if a customer comes, he lands on to a node and remains there forever. This has two major issues: 1) We have to overprovision each node with ability to handle worse case capacity. 2) If two or three high profile customers lands on to same node then we need to move them manually. 3) We need to cut over new nodes and we already have over 100+ nodes.  Its a pain managing these nodes and I waste lot of my time in chasing node specific issues. I loath when I know I have to chase this env issue. I really hate human intervention as if it were up to me I would just automate thing and just enjoy the fruits of automation and spend quality time on major issues rather than mundane task,call me lazy but thats a good quality. So Finally now I am at a stage where I can put nodes behing HAProxy in QA env. today we were testing the HA config and first problem I immediat...

Adding Jitter to cache layer

Thundering herd is an issue common to webapp that rely on heavy caching where if lots of items expire at the same time due to a server restart or temporal event, then suddenly lots of calls will go to database at same time. This can even bring down the database in extreme cases. I wont go into much detail but the app need to do two things solve this issue. 1) Add consistent hashing to cache layer : This way when a memcache server is added/removed from the pool, entire cache is not invalidated.  We use memcahe from both python and Java layer and I still have to find a consistent caching solution that is portable across both languages. hash_ring and spymemcached both use different points for server so need to read/test more. 2) Add a jitter to cache or randomise the expiry time: We expire long term cache  records every 8 hours after that key was added and short term cache expiry is 2 hours. As our customers usually comes to work in morning and access the cloud file server it ...

Spring 3.2 quartz 2.1 Jobs added with no trigger must be durable.

I am trying to enable HA on nodes and in that process I found that in a two test node setup a job that has a frequency of 10 sec was running into deadlock. So I tried upgrading from Quartz 1.8 to 2.1 by following the migration guide but I ran into an exception that says "Jobs added with no trigger must be durable.". After looking into spring and Quartz code I figured out that now Quartz is more strict and earlier the scheduler.addJob had a replace parameter which if passed to true would skip the durable check, in latest quartz this is fixed but spring hasnt caught up to this. So what do you do, well I jsut inherited the factory and set durability to true and use that public class DurableJobDetailFactoryBean extends JobDetailFactoryBean {     public DurableJobDetailFactoryBean() {         setDurability(true);     } } and used this instead of JobDetailFactoryBean in the spring bean definition     <bean i...