Hadoop Adventures At Spotify Adam Kawa Data Engineer
How many times were you told bad news?
How many times did your excellent weekend change into a workday?
How many times did you change your passion for glory?
How many times did you surrender?
Many typos and grammar errors ;) Hadoop nightmares guaranteed!
My answer to each of these questions: At least once
Why Hadoop Adventures? A rapidly growing cluster! The disk space is consumed quicker and quicker Configuration settings stop working Masters have troubles coordinating slaves More data + users + jobs Own mistakes Many nodes were added
What feelings did you get when it happened? Responsibility, Critical, Expensive, Interruptions, Frustration, Stress, Email, Ashamed, Cooperation, Learning, Surprise, Fun, Exciting, Happy Pride
About Me I needed to learn more about operating systems, hardware, networking... A Java developer who was developing MapReduce, Pig and Hive jobs who started operating Hadoop and HBase on a 5-node cluster who later joined Spotify to meet a 138x larger Elephant
About The Elephant Hands off a little boy! I met this Elephant He grows rapidly from 60 to 690 nodes He stores 6.29PT of data He receives 50TB of data each day He seems to be the largest in Europe He did not like me at al His friends did not like me too We had problems living together! I want to make them PUBLIC now!
What Will I Talk About? Five real (and my favourite) Hadoop incidents that broke our cluster or made it very unstable
What Will I Share? Real problems Real mistakes Real lessons learned Real graphs Real numbers Real emails Real excerpts from our conversations
What Will Also I Share? The mistake that I made and I did not like to talk about Feeling: Ashamed
Adventure 1 (February 2013) Troubles running more resource-intensive Hive queries
Problem A user can not run resource-intensive Hive queries It happened immediately after significantly expanding the cluster
Description The queries are valid The queries run successfully on small datasets But they fail on large datasets Surprisingly they run successfully through other user accounts The user has right permissions to HDFS directories and Hive tables
Observations When this user runs a resource-intensive query The cluster experiences stability problems The NameNode becomes less responsive
It was losing the connection with DataNodes and mark them dead
But the DataNode daemons are running completely fine
The NameNode is throwing thousands of warnings and exceptions 14592 times only during 8 min (4768/min in a peak)
Normally Hadoop is a very trusty elephant The username comes from the client machine (and is not verified) The groupname is resolved on the NameNode server Using the shell command ''id -Gn <username>'' If a user does not have an account on the NameNode server The ExitCodeException exception is thrown
Possible Fixes Create an user account on the NameNode server (dirty, insecure) Use AD/LDAP for a user-group resolution hadoop.security.group.mapping.ldap.* settings If you also need the full-authentication, deploy Kerberos
Our Fix We decided to use LDAP for a user-group resolution However, LDAP settings in Hadoop did not work for us
Because posixGroup is not a supported filter group class We found a workaround using nsswitch.conf
Lesson Learned Know who is going to use your cluster Know who is abusing the cluster (HDFS access and MapReduce jobs) Parse the NameNode logs regularly Look for FATAL, ERROR, Exception messages Especially before and after expanding the cluster
Adventure 2 (March 2013) DataNodes become blocked sometimes (Part 1)
Problem The DataNodes are marked dead by the NameNode The DataNodes have not sent a heartbeat for a long time Last Contact (seconds)
When a DataNode Is Marked Dead A costly block replication process is started Alive DataNodes consume resources to recreate missing replicas Map tasks are processing non-local blocks Tasks and jobs experience problems They can fail or be not started at all (BlockMissingException)
Observations $ps shows the DataNode daemons exist on their servers sudo jstack -F <datanode-pid> does not show anything interesting Some of the DataNodes automatically re-join the cluster The DataNodes seem to be blocked for a while
First (Relief) Idea! Let's temporarily increase the datanode-liveness interval to get some breathing room. . Wouter: Maybe we should up the
dfs.namenode.heartbeat.recheck-interval? Adam: What value could be OK? 15, maybe 20 minutes? Wouter: 20 maybe? It will give us a bit more breathing room. Adam: Yes, the replication will not start so quickly..
A patch was quickly written, reviewed and deployed. The formula is: 2 * dfs.namenode.heartbeat.recheck.interval + 10 * dfs.heartbeat.interval and it should give us 2 * 600 sec + 10 * 3 sec → 20min:30sec
Crazy Idea! Let's do other (important) tasks after deploying this (simple) change … and do not measure its impact
Adventure 3 (March 2013) The cluster becomes unstable after around 1 hour of uptime
Problem Each time, around 1 hour after restarting the NameNode Majority (or all) of the DataNodes are marked dead by the NameNode $ps shows the the DataNode daemons The DataNode servers are running fine
Question? Why does it happen each time, exactly around 1 hour after restarting the NameNode?
First Idea! ''When a DataNode initial y starts up, as well as every hour thereafter, it sends a block report to the NameNode'' You can read it in * a book, * blog posts, * even see it in the code!
Maybe? A storm of block reports coming after the first hour heavily overloads the NameNode? starts a heavy garbage collection phase that freezes the NameNode?
Could it be a right fix? Increase dfs.blockreport.initialDelay to something bigger than zero Delay for first block report in seconds – a random value between 0 and dfs.blockreport.initialDelay.
Because Changing initialDelay requires a restart of the NameNode The restart will be longer (the block reports are sent with a delay) … and it is a sunny Sunday afternoon Let's also deploy memory+GC changes to the NN at the same time We will solve more problems in fewer iterations!
A couple of hours later...
Wouter: Yeah, looks like we're pretty good! Adam: Good timing, one evening before Monday... Johannes: Yeah, maybe I'll get two or three hours of weekend!
A couple of days later...
WARNING : There are 10779564 missing blocks. Please check the logs or run fsck in order to identify the missing blocks. Number of dead DataNodes
A Mail Of Shame! I had a dilemma whether to silently fix this interval or send this email...
A Reply Of Support!
Next (Shocking) Finding! NO storm of block report after the 1st hour means that ... tuning heap and GC helped
Too Many Lessons Learned Measure the impact of each single change Never make bulk changes to the cluster Double-check a description of a configuration parameter Double-check the default values of configuration parameters Question (almost) everything Troubleshoot the cluster together interactively and non-interactively Share the knowledge, even if you make a mistake Give a support to your team-mates, even if they fail
Adventure 2 (March 2013) DataNodes become blocked sometimes (Part 2)
Observations sudo jstack -F <datanode-pid> does not show anything interesting sudo -u hdfs jstack <datanode-pid> shows something interesting 462 threads are blocked/waiting 19 threads are runnable The threads are waiting for the same lock, to finish operations like: org.apache.hadoop.hdfs.server.datanode.DataXceiver.readBlock org.apache.hadoop.hdfs.server.datanode.fsdataset.impl.FsDatasetImpl.createRbw org.apache.hadoop.hdfs.server.datanode.fsdataset.impl.FsVolumeList.getDfsUsed org.apache.hadoop.hdfs.server.datanode.fsdataset.impl.FsDatasetImpl.initReplicaRecovery org.apache.hadoop.hdfs.server.datanode.fsdataset.impl.FsDatasetImpl.finalizeBlock
The lock is held by other thread that is … waiting for something else
recoverRbw tries to stop the other thread it has the lock (the monitor object) that the other is waiting for This deadlock is Described in HDFS-3655 Together Duplicated by HDFS-5016 with many Related to HDFS-4851 other threads
Lesson Learned Master the troubleshooting and monitoring tools Contribute to Apache Hadoop more Send a post to Hadoop mailing list Raise a well-documented JIRA ticket Implement a patch
Adventure 4 (March 2013) Tasks in Hive jobs are constantly ''KILLED UNCLEAN by the user''
Problem Surprisingly, ad-hoc Hive queries are running extremely long Thousands of task are constantly being killed
Only 1 task failed, 2x more task were killed than were completed
The logs show that the JobTracker gets a request to kil the tasks Who can actually send a kill request? User (using e.g. mapred job -kill-task) JobTracker (a speculative duplicate, or when a whole job fails) Fair Scheduler (if the preemption is enabled)
Two suspicious characters have a good alibi Users She/he is friendly, patient and respective to others JobTracker The speculative execution is disabled Jobs have not fail (yet)
Key Observations Tasks are usually killed quickly after the start Surviving tasks are running fine for long time Ad-hoc Hive queries are running in their own Fair Scheduler's pool
Eureka! FairScheduler prefers to kil young! Preempt the newest tasks in an over-share pools to forcibly give their slots to starving pools
Hive pool was running over its minimum and fair shares Other pools were starving So that Fair Scheduler was (legal y) killing Hive tasks from time to time Fair Scheduler can kill to be KIND...
Possible Fixes Tune minimum shares based on your workload Tune preemption timeouts based on your workload Disable the preemption Limit the number of map/reduce tasks in a pool Limit the number of jobs in a pool Alternative Capacity Scheduler
Lessons Learned A scheduler should NOT be considered as a ''black-box''
Adventure 5 (April 2013) JobTracker runs super slowly
Problem JobTracker becomes super slow... (but used to be super snappy) Everybody is annoyed by unresponsive JobTracker web interface
Observations Many large jobs are running on the cluster Only the largest one (at that time) run 58.538 tasks It aims to process 21.82 TB of data 1.5 year of data from one of our datasets It is an ad-hoc Hive job It is still easy to run a large Hive job even if hive.mapred.mode is set to strict...
Possible Solution Limit the number of tasks per job using mapred.jobtracker.maxtasks.per.job Unfortunately, there are no separate properties for map and reduce tasks
Real-Life Dialogue Adam: What value for mapred.jobtracker.maxtasks.per.job? Sven: Maybe 50K Adam: Why not 30K? Sven: We can run large production jobs and it is safer to start with a high value and maybe decrease it later. Adam: Hmmm? Sounds good...!
But maybe 40K, not 50K? ;) Sven: OK. Let's meet in the middle...
The previous approach is a great example of “guesstimation” Negotiation skills are required! Question? Should a real data-driven company make such a decision based on a guess or data?
The data and the answer Based on 436K jobs from two months The largest production job created 22.6K tasks (still to many!) The jobs that create more than 23K tasks Usual y ad-hoc Hive queries or Python streaming jobs These jobs fail, or are killed by impatient users in most cases Hadoop can be used to … process data generated by Hadoop ;)
Lessons Learned Make data-driven decisions Use Hadoop to analyze … Hadoop Administrators should cooperate with developers+analysts Developers+analysts will often adapt to the changes Negotiation skills are stil useful
A technical blog
with some bytes of humour! More Adventures at HakunaMapData.com Nodes are marked dead due to a heavy swapping The out-of-memory killer is killing TaskTrackers The script for daily data cleaning runs more than one day!
Adventure The nearest future will be BIGGER YARN + Tez + Fast SQL
Grand Lesson Learned Hadoop is like a kid! It needs love and care It can make you proud, but it also will cause problems It wil grow and change quickly It wil bring new friends home It wil surprise you
Are there any issues questions? ;)
One Question: What can happen after some time of simultaneous development of MapReduce jobs maintenance of a large cluster and listening to perfect music for every moment?
A Possible Answer: You may discover Hadoop in the lyrics of many popular songs!