Skip to main content

Move fast break things but with monitoring

We run a complex system with multiple services and every 2 or 3 week we  update the Java applications.  I want to do it every week as most applications are stateless and can be patched anytime but the application serving the main website is using sticky session. We are working to make it failover sessions, once we do that, we can do mid week deployment and that will allow us to go faster than 3 weeks.  This week I pushed a huge infrastructure change related to user Id generation. I had asked ops team to check the status of new relic after the midnight deployment and it looked like this so everyone was happy.


I woke up and checked new relic mobile app and things looked ok to me. After finishing my morning routines I ran my daily exception report and one thing that caught the eye was 90K exceptions in last 12 hours in one of the files I had changed.  To gauge the impact I went in new relic and it showed me an error rate of 0.07 in one of the app

I then checked new relic and I see this blip that caused 90k errors, when the blip was there the status must have been red but then it was quickly green and Ops team didnt caught it
 


The issue was due to a wrong method invocation in one of the class used by this one specific application and it took just 15 min to fix after reproducing with a testcase.  So why didnt QA/UAT/automated tests caught it, well the issue was like a heisen bug and would occur only when the object is missing in cache. 99% of the calls would go to cache and only 0.07% were going to dao layer that had the bug.  I quickly made a patch and ops deployed it and I can see the issue is now gone.

Had there been no monitoring it would have been difficult to catch these kinds of bugs. New relic and internal monitoring tools makes life easy as it exposes anomalies.

Comments

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...
Post a Comment
Read more

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 ...
Post a Comment
Read more

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...
7 comments
Read more