Skip to main content

Deal with corrupted messages in Apache Kafka

Image
By
Listen:

Under some strange circumstances, it can happen that a message in a Kafka topic is corrupted. This often happens when using 3rd party frameworks with Kafka. In addition, Kafka < 0.9 does not have a lock on Log.read() at the consumer read level, but does have a lock on Log.write(). This can lead to a rare race condition as described in KAKFA-2477 [1]. A likely log entry looks like this:

ERROR Error processing message, stopping consumer: (kafka.tools.ConsoleConsumer$) kafka.message.InvalidMessageException: Message is corrupt (stored crc = xxxxxxxxxx, computed crc = yyyyyyyyyy

Kafka-Tools

Kafka stores the offset of each consumer in Zookeeper. To read the offsets, Kafka provides handy tools [2]. But you can also use zkCli.sh, at least to display the consumer and the stored offsets. First we need to find the consumer for a topic (> Kafka 0.9):

bin/kafka-consumer-groups.sh --zookeeper management01:2181 --describe --group test

Prior to Kafka 0.9, the only way to get this information was to use zkCli.sh (or similar tools) to find the consumer group. Since debugging with zkCli is a bit frustrating, I personally use Yahoo's kafka-manager [3]. 

Let's assume the consumers are stored in Zookeeper under /consumer, the command to find the offset looks like:

ls /consumer/test/offsets
[1]
get /consumer/test/offsets/1
[15]

With Kafka that command would look like:

bin/kafka-run-class.sh kafka.tools.ConsumerOffsetChecker --group console-1 --zookeeper zknode1:2181

Group     Topic   Pid   Offset   logSize   Lag   Owner
console-1 test    1     15       337       326   none

Once the offset is found, it can be incremented to force the consumer to read the next available message. Before doing this, Kafka must be shut down. 

bin/kafka-run-class.sh kafka.tools.UpdateOffsetsInZK latest 16 test

After the restart, Kafka should be able to read the next message, unless that message is also corrupted. And yes, the corrupted message is lost and cannot be recovered, so it's always a good idea to implement a CRC check before any message gets to Kafka.

A code-based approach is also available [4]. This involves creating a subclass of the ConsumerIterator that catches the message exception, replaces it with a dummy message, and moves on to the next message. Of course, the corrupted message is also lost in this case.

Labels:
Location: Austin, TX, USA

Comments

Post a Comment

Popular posts from this blog

Hive query shows ERROR "too many counters"

By Anonymous
A hive job face the odd " Too many counters:"  like Ended Job = job_xxxxxx with exception 'org.apache.hadoop.mapreduce.counters.LimitExceededException(Too many counters: 201 max=200)' FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.MapRedTask Intercepting System.exit(1) These happens when operators are used in queries ( Hive Operators ). Hive creates 4 counters per operator, max upto 1000, plus a few additional counters like file read/write, partitions and tables. Hence the number of counter required is going to be dependent upon the query.  To avoid such exception, configure " mapreduce.job.counters.max " in mapreduce-site.xml to a value above 1000. Hive will fail when he is hitting the 1k counts, but other MR jobs not. A number around 1120 should be a good choice. Using " EXPLAIN EXTENDED " and " grep -ri operators | wc -l " print out the used numbers of operators. Use this value to tweak the MR s
2 comments
Read more

AI's False Reality: Understanding Hallucination

By
Artificial Intelligence (AI) has leapfrogged to the poster child of technological innovation, on track to transform industries in a scale similar to the Industrial Revolution of the 1800s. But in this case, as cutting-edge technology, AI presents its own unique challenge, exploiting our human behavior of "love to trust", we as humans face a challenge: AI hallucinations. This phenomenon, where AI models generate outputs that are factually incorrect, misleading, or entirely fabricated, raises complex questions about the reliability and trust of AI models and larger systems. The tendency for AI to hallucinate comes from several interrelated factors. Overfitting – a condition where models become overly specialized to their training data – can lead to confident but wildly inaccurate responses when presented with novel scenarios (Guo et al., 2017). Moreover, biases embedded within datasets shape the models' understanding of the world; if these datasets are flawed or unreprese
Post a Comment
Read more