We right now have an existing running flink job which contains keyed states whose max parallelism is set to 128. As our data grows, we are concerned that 128 is not enough any more in the future. I want to know if we have a way to change the max parallelism by modifying the savepoint? Or is there any way to do that?
You can use the State Processor API to accomplish this. You will read the state from a savepoint taken from the current job, and write that state into a new savepoint with increased max parallelism. https://nightlies.apache.org/flink/flink-docs-stable/docs/libs/state_processor_api/
Your job should perform well if the maximum parallelism is (roughly) 4-5 times the actual parallelism. When the max parallelism is only somewhat higher than the actual parallelism, then you have some slots processing data from just one key group, and others handling two key groups, and that imbalance wastes resources.
But going unnecessarily high will exact a performance penalty if you are using the heap-based state backend. That's why the default is only 128, and why you don't want to set it to an extremely large value.
Related
My question is about knowing a good choice for parallelism for operators in a flink job in a fixed cluster setting. Suppose, we have a flink job DAG containing map and reduce type operators with pipelined edges between them (no blocking edge). An example DAG is as follows:
Scan -> Keyword Search -> Aggregation
Assume a fixed size cluster of M machines with C cores each and the DAG is the only workflow to be run on the cluster. Flink allows the user to set the parallelism for individual operators. I usually set M*C parallelism for each operator. But is this the best choice from performance perspective (e.g. execution time)? Can we leverage the properties of the operators to make a better choice? For example, if we know that aggregation is more expensive, should we assign M*C parallelism to only the aggregation operator and reduce the parallelism for other operators? This hopefully will reduce the chances of backpressure too.
I am not looking for a proper formula that will give me the "best" parallelism. I am just looking for some kind of an intuition/guideline/ideas that can be used to make a decision. Surprisingly, I could not find much literature to read on this topic.
Note: I am aware of the dynamic scaling reactive mode in recent Flink. But my question is about a fixed cluster with only one workflow running, which means that the dynamic scaling is not relevant. I looked at this question, but did not get an answer.
I think about this a little differently. From my perspective, there are two key questions to consider:
(1) Do I want to keep the slots uniform? Or in other words, will each slot have an instance of every task, or do I want to adjust the parallelism of specific tasks?
(2) How many cores per slot?
My answer to (1) defaults to "keep things uniform". I haven't seen very many situations where tuning the parallelism of individual operators (or tasks) has proven to be worthwhile.
Changing the parallelism is usually counterproductive if it means breaking an operator chain. Doing it where's a shuffle anyway can make sense in unusual circumstances, but in general I don't see the point. Since some of the slots will have instances of every operator, and the slots are all uniform, why is it going to be helpful to have some slots with fewer tasks assigned to them? (Here I'm assuming you aren't interested in going to the trouble of setting up slot sharing groups, which of course one could do.) Going down this path can make things more complex from an operational perspective, and for little gain. Better, in my opinion, to optimize elsewhere (e.g., serialization).
As for cores per slot, many jobs benefit from having 2 cores per slot, and for some complex jobs with lots of tasks you'll want to go even higher. So I think in terms of an overall parallelism of M*C for simple ETL jobs, and M*C/2 (or lower) for jobs doing something more intense.
To illustrate the extremes:
A simple ETL job might be something like
source -> map -> sink
where all of the connections are forwarding connections. Since there is only one task, and because Flink only uses one thread per task, in this case we are only using one thread per slot. So allocating anything more than one core per slot is a complete waste. And the task is probably i/o bound anyway.
At the other extreme, I've seen jobs that involve ~30 joins, the evaluation of one or more ML models, plus windowed aggregations, etc. You certainly want more than one CPU core handling each parallel slice of a job like that (and more than two, for that matter).
Typically most of the CPU effort goes into serialization and deserialization, especially with RocksDB. I would try to figure out, for every event, how many RocksDB state accesses, keyBy's, and rebalances are involved -- and provide enough cores that all of that ser/de can happen concurrently (if you care about maximizing throughput). For the simplest of jobs, one core can keep up. By the time to you get to something like a windowed join you may already be pushing the limits of what one core can keep up with -- depending on how fast your sources and sinks can go, and how careful you are not to waste resources.
Example: imagine you are choosing between a parallelism of 50 with 2 cores per slot, or a parallelism of 100 with 1 core per slot. In both cases the same resources are available -- which will perform better?
I would expect fewer slots with more cores per slot to perform somewhat better, in general, provided there are enough tasks/threads per slot to keep both cores busy (if the whole pipeline fits into one task this might not be true, though deserializers can also run in their own thread). With fewer slots you'll have more keys and key groups per slot, which will help to avoid data skew, and with fewer tasks, checkpointing (if enabled) will be a bit better behaved. Inter-process communication is also a little more likely to be able to take an optimized (in-memory) path.
we are trying to setup a Flink stateful job using RocksDB backend.
We are using session window, with 30mins gap. We use aggregateFunction, so not using any Flink state variables.
With sampling, we have less than 20k events/s, 20 - 30 new sessions/s. Our session basically gather all the events. the size of the session accumulator would go up along time.
We are using 10G memory in total with Flink 1.9, 128 containers.
Following's the settings:
state.backend: rocksdb
state.checkpoints.dir: hdfs://nameservice0/myjob/path
state.backend.rocksdb.memory.managed: true
state.backend.incremental: true
state.backend.rocksdb.memory.write-buffer-ratio: 0.4
state.backend.rocksdb.memory.high-prio-pool-ratio: 0.1
containerized.heap-cutoff-ratio: 0.45
taskmanager.network.memory.fraction: 0.5
taskmanager.network.memory.min: 512mb
taskmanager.network.memory.max: 2560mb
From our monitoring of a given time,
rocksdb memtable size is less than 10m,
Our heap usage is less than 1G, but our direct memory usage (network buffer) is using 2.5G. The buffer pool/ buffer usage metrics are all at 1 (full).
Our checkpoints keep failing,
I wonder if it's normal that the network buffer part could use up this much memory?
I'd really appreciate if you can give some suggestions:)
Thank you!
For what it's worth, session windows do use Flink state internally. (So do most sources and sinks.) Depending on how you are gathering the session events into the session accumulator, this could be a performance problem. If you need to gather all of the events together, why are you doing this with an AggregateFunction, rather than having Flink do this for you?
For the best windowing performance, you want to use a ReduceFunction or an AggregateFunction that incrementally reduces/aggregates the window, keeping only a small bit of state that will ultimately be the result of the window. If, on the other hand, you use only a ProcessWindowFunction without pre-aggregation, then Flink will internally use an appending list state object that when used with RocksDB is very efficient -- it only has to serialize each event to append it to the end of the list. When the window is ultimately triggered, the list is delivered to you as an Iterable that is deserialized in chunks. On the other hand, if you roll your own solution with an AggregateFunction, you may have RocksDB deserializing and reserializing the accumulator on every access/update. This can become very expensive, and may explain why the checkpoints are failing.
Another interesting fact you've shared is that the buffer pool / buffer usage metrics show that they are fully utilized. This is an indication of significant backpressure, which in turn would explain why the checkpoints are failing. Checkpointing relies on the checkpoint barriers being able to traverse the entire execution graph, checkpointing each operator as they go, and completing a full sweep of the job before timing out. With backpressure, this can fail.
The most common cause of backpressure is under-provisioning -- or in other words, overwhelming the cluster. The network buffer pools become fully utilized because the operators can't keep up. The answer is not to increase buffering, but to remove/fix the bottleneck.
I'm trying to understand what are the important features I need to take into consideration before submitting a Flink job.
My question is what is the number of parallelism, is there an upper bound(physically)? and how can the parallelism impact the performance of my job?
For example, I have a CEP Flink job that detects a pattern from unkeyed Stream, the number of parallelism will always be 1 unless I partition the datastream with KeyBy operator.
Plz Correct me if I'm wrong :
If I partition the data stream, then I will have a number of parallelism equals to the number of different keys. but the problem is that the pattern matching is being done independently for each key so I can't define a pattern that requires information from 2 partitions that have different keys.
It's not bad to use Flink with parallelism = 1. But it defeats the main purpose of using Flink (being able to scale).
In general, you should not have a higher parallelism than your cores (physical or virtual depends on the use case) as you want to saturate your cores as much as possible. Anything over that will negatively impact your performance as it requires more communication overhead and context switching. By scaling out, you can add cores from distributed compute nodes in a network, which is the main benefit of using big data technologies vs. writing application by hand.
As you said you can only use the parallelism if you partition your data. If you have an algorithm that needs all data, you need to process it on one core eventually. However, usually you can do lots of preprocessing (filtering, transformation) and partial aggregations in parallel before combining the data at a final core. For example, think of simply counting all events. You can count the data of each partition and then simply sum up the partial counts in a final step, which scales almost perfectly.
If your algorithm does not allow splitting it up, then your use case may not allow distributed processing. In that case, Flink is not a good fit. However, it's worth exploring if alternative algorithms (sometimes approximate) would suffice your use case as well. That's the art of data engineering to split monolithic algorithms into parallelizable sub-algorithms.
I am pretty new to flink and about to load our first production version. We have a stream of data. The stateful filter is checking if the data is new.
would it be better to split the stream to different jobs to gain more control on the parallelism as shown in option 1 or option 2 is better ?
following the documentation recommendation. should I put uid per operator e.g :
dataStream
.uid("firstid")
.keyBy(0)
.flatMap(flatMapFunction)
.uid("mappedId)
should I add rebalance after each uid if at all?
what is the difference if I setMaxParallelism as described here or setting parallelism from flink UI/cli ?
You only need to define .uid("someName") for your stateful operators. Not much need for operators which do not hold state as there is nothing in the savepoints that needs to be mapped back to them (more on this here). Won't hurt if you do though.
rebalance will only help you in the presence of data skew and that only if you aren't using keyed streams. If you process data based on a key, and your load isn't uniformly distributed across your keys (ie you have loads of "hot" keys) then rebalancing won't help you much.
In your example above I would start Option 2 and potentially move to Option 1 if the job proves to be too heavy. In general stateless processes are very fast in Flink so unless you want to add other consumers to the output of your stateful filter then don't bother to split it up at this stage.
There isn't right and wrong though, depends on your problem. Start simple and take it from there.
[Update] Re 4, setMaxParallelism if I am not mistaken defines the number of key groups and thus the maximum number of parallel instances your stream can be rescaled to. This is used by Flink internally but it doesn't set the parallelism of your job. You usually have to set that to some multiple of the actually parallelism you set for you job (via -p <n> in the CLI/UI when you deploy it).
I am planning to decide on how many nodes should be present on Kafka Cluster. I am not sure about the parameters to take into consideration. I am sure it has to be >=3 (with replication factor of 2 and failure tolerance of 1 node).
Can someone tell me what parameters should be kept in mind while deciding the cluster size and how they effect the size.
I know of following factors but don't know how it quantitatively effects the cluster size. I know how it qualitatively effect the cluster size. Is there any other parameter which effects cluster size?
1. Replication factor (cluster size >= replication factor)
2. Node failure tolerance. (cluster size >= node-failure + 1)
What should be cluster size for following scenario while consideration of all the parameters
1. There are 3 topics.
2. Each topic has messages of different size. Message size range is 10 to 500kb. Average message size being 50kb.
3. Each topic has different partitions. Partitions are 10, 100, 500
4. Retention period is 7 days
5. There are 100 million messages which gets posted every day for each topic.
Can someone please point me to relevant documentation or any other blog which may discuss this. I have google searched it but to no avail
As I understand, getting good throughput from Kafka doesn't depend only on the cluster size; there are others configurations which need to be considered as well. I will try to share as much as I can.
Kafka's throughput is supposed to be linearly scalabale with the numbers of disk you have. The new multiple data directories feature introduced in Kafka 0.8 allows Kafka's topics to have different partitions on different machines. As the partition number increases greatly, so do the chances that the leader election process will be slower, also effecting consumer rebalancing. This is something to consider, and could be a bottleneck.
Another key thing could be the disk flush rate. As Kafka always immediately writes all data to the filesystem, the more often data is flushed to disk, the more "seek-bound" Kafka will be, and the lower the throughput. Again a very low flush rate might lead to different problems, as in that case the amount of data to be flushed will be large. So providing an exact figure is not very practical and I think that is the reason you couldn't find such direct answer in the Kafka documentation.
There will be other factors too. For example the consumer's fetch size, compressions, batch size for asynchronous producers, socket buffer sizes etc.
Hardware & OS will also play a key role in this as using Kafka in a Linux based environment is advisable due to its pageCache mechanism for writing data to the disk. Read more on this here
You might also want to take a look at how OS flush behavior play a key role into consideration before you actually tune it to fit your needs. I believe it is key to understand the design philosophy, which makes it so effective in terms of throughput and fault-tolerance.
Some more resource I find useful to dig in
https://engineering.linkedin.com/kafka/benchmarking-apache-kafka-2-million-writes-second-three-cheap-machines
http://blog.liveramp.com/2013/04/08/kafka-0-8-producer-performance-2/
https://grey-boundary.io/load-testing-apache-kafka-on-aws/
https://cwiki.apache.org/confluence/display/KAFKA/Performance+testing
I had recently worked with kafka and these are my observations.
Each topic is divided into partitions and all the partitions of a topic are distributed across kafka brokers; first of all these help to save topics whose size is larger than the capacity of a single kafka broker and also they increase the consumer parallelism.
To increase the reliability and fault tolerance,replications of the partitions are made and they do not increase the consumer parallelism.The thumb rule is a single broker can host only a single replica per partition. Hence Number of brokers must be >= No of replicas
All partitions are spread across all the available brokers,number of partitions can be irrespective of number of brokers but number of partitions must be equal to the number of consumer threads in a consumer group(to get best throughput)
The cluster size should be decided keeping in mind the throughput you want to achieve at consumer.
The total MB/s per broker would be:
Data/Day = (100×10^6 Messages / Day ) × 0.5MB = 5TB/Day per Topic
That gives us ~58MB/s per Broker. Assuming that the messages are equally split between partitions, for the total cluster we get: 58MB/s x 3 Topics = 178MB/s for all the cluster.
Now, for the replication, you have: 1 extra replica per topic. Therefore this becomes 58MB/sec/broker INCOMING original data + 58MB/sec/broker OUTGOING replication data + 58MB/sec/broker INCOMING replication data.
This gets about ~136MB/s per broker ingress and 58MB/s per broker egress.
The systems load will get very high and this is without taking into consideration any stream processing.
The system load could be handled by increasing the number of brokers and splitting your topics to more specific partitions.
If your data are very important, then you may want a different (high) replication factor. Fault tolerance is also an important factor for deciding the replication.
For example, if you had very very important data, apart from the N active brokers (with the replicas) that are managing your partitions, you may require to add stand-by followers in different areas.
If you require very low latency, then you may want to further increase your partitions (by adding additional keys). The more keys you have, the fewer messages you will have on each partition.
For low latency, you may want a new cluster (with the replicas) that manages only that special topic and no additional computation is done to other topics.
If a topic is not very important, then you may want to lower the replication factor of that particular topic and be more elastic to some data loss.
When building a Kafka cluster, the machines supporting your infrastructure should be equally capable. That is since the partitioning is done with round-robin style, you expect that each broker is capable of handling the same load, therefore the size of your messages does not matter.
The load from stream processing will also have a direct impact. A good software to manage your kafka monitor and manage your streams is Lenses, which I personally favor a lot since it does an amazing work with processing real-time streams