I need to build a server that reads large csv data files (100GBs) in a directory, transforms some fields and streams them to a Hadoop cluster.
These files are copied over from other servers at random time (100s times/day). It takes a long time to finish copying a file.
I need to:
Regularly check for new files to process (i.e., encrypt and stream)
Check if a csv is completely copied over to kick off encryption
Process Stream multiple files in parallel, but prevent two processes
to stream the same file
Mark files being streamed successfully
Mark
files being streamed unsuccessfully and restart the streaming
process.
My question is: is there an open source ETL tool that provide all of the 5, and works well with Hadoop/Spark Stream? I assume this process is fairly standard, but I couldn't find any yet.
Thank you.
Flume or Kafka will serve your purpose. Both are well integrated with Spark and Hadoop.
Try taking a look at the great library https://github.com/twitter/scalding. Maybe it can point you in the right direction :)
Related
I have an issue with small files and HDFS.
Scenario: I am using NiFi to read messages from the Kafka topic, these are all really small.
Requirement: to store these raw messages of data in HDFS(for replay capability)...before doing further processing on them.
I was thinking using Hadoop Archive (HAR) on them periodically. Is that something i can do through NiFi? the har command seems like a command line thing rather than something that i could execute through Nifi? Would love to know a solution that can achieve my requirement, without bringing down HDFS due to the small files.
Ginil
You can execute command line inside Nifi with ExecuteProcess processor :
http://nifi.apache.org/docs/nifi-docs/components/org.apache.nifi/nifi-standard-nar/1.6.0/org.apache.nifi.processors.standard.ExecuteProcess/
You can also take a look at Kafka-connect HDFS for putting kafka records into HDFS.
I don't know how to build architecture for following use case:
I have an Web application where users can upload files(pdf&pptx) and directories to be processed. After upload is complete web application put this files and directories in HDFS, then send a messages on kafka with path to this files.
Spark Application read messages from kafka streaming, collect them on master(driver), and after that process them. I collect messages first because i need to move the code to data, and not move data where the message is received. I understood that spark assign job to executor which already have file locally.
I have issues with kafka because i was forced to collect them first for the above reason, and when want to create checkpoint app crash "because you are attempting to reference SparkContext from a broadcast variable" even if the code run before adding checkpointing( I use sparkContext there because i need to save data to ElasticSearch and PostgreSQL. I don't know how exactly i can do code upgrading in this conditions.
I read about hadoop small files problems, and I understand what problems are in this case. I read that HBase is a better solution to save small files than just save in hdfs. Other problem in hadoop small files problems is big number of mappers and reducers created for computation, but i don't understand if this problem there in spark.
What is the best architecture for this use case?
How to do Job Scheduling? It's kafka good for that? or I need to use other service like rabbitMQ or something else?
Exist some method to add jobs to an running Spark application through some REST API?
How is the best way to save files? Is better to use Hbase because i have small files(<100MB)? Or I need to use SequenceFile? I think SequenceFile isn't for my use case because i need to reprocess some files randomly.
What is the best architecture do you think for this use case?
Thanks!
There is no one single "the best" way to build architecture. You need to make decisions and stick to them. Make the architecture flexible and decoupled so that you can easily replace components if needed.
Consider following stages/layers in your architecture:
Retrieval/Acquisition/Transport of source data (files)
Data processing/transformation
Data archival
As a retrieval component, I would use Flume. It is flexible, supports a lot of sources, channels (including Kafka) and sinks. In your case you can configure source that monitors the directory and extracts the newly received files.
For data processing/transformation - it depends what task you are solving. You probably decided on Spark Streaming. Spark streaming can be integrated with Flume sink (http://spark.apache.org/docs/latest/streaming-flume-integration.html) There are other options available, e.g. Apache Storm. Flume combines very well with Storm. Some transformations can also be applied in Flume.
For data archival - do not store/archive the files directly in Hadoop, unless they are bigger than few hundredths of megabytes. One solution would be to put them in HBase.
Make your architecture more flexible. I would place processed files in a temporary HDFS location and have some job regualarly archive them into zip, HBase, Hadoop Archive (there is such an animal) or any other solution.
Consider using Apache NiFi (aka HDF - Hortonworks Data Flow). It uses internally queues, provides a lot of processors. It can make your life easier and get the workflow developed in minutes. Give it a try. There is nice Hortonworks tutorial which , combined with HDP Sandbox running on a virtual machine/Docker, can bring you up to speed in very short time (1-2 hours?).
I have created a real time application in which I am writing data streams to hdfs from weblogs using flume, and then processing that data using spark stream. But while flume is writing and creating new files in hdfs spark stream is unable to process those files. If I am putting the files to hdfs directory using put command spark stream is able to read and process the files. Any help regarding the same will be great.
You have detected the problem yourself: while the stream of data continues, the HDFS file is "locked" and can not be read by any other process. On the contrary, as you have experienced, if you put a batch of data (that's yur file, a batch, not a stream), once it is uploaded it is ready for being read.
Anyway, and not being an expert on Spark streaming, it seems from the Spark Streaming Programming Guide, Overview section, that you are not performing the right deployment. I mean, from the picture shown there, it seems the streaming (in this case generated by Flume) must be directly sent to Spark Streaming engine; then the results will be put in HDFS.
Nevertheless, if you want to maintain your deployment, i.e. Flume -> HDFS -> Spark, then my suggestion is to create mini-batches of data in temporal HDFS folders, and once the mini-batches are ready, store new data in a second minibatch, passing the first batch to Spark for analysis.
HTH
In addition to frb's answer: which is correct - SparkStreaming with Flume acts as an Avro RPC Server - you'll need to configure an AvroSink which points to your SparkStreaming instance.
with spark2, now you can connect directly your spark streaming to flume, see official docs, and then write once on HDFS at the end of the process.
import org.apache.spark.streaming.flume._
val flumeStream = FlumeUtils.createStream(streamingContext, [chosen machine's hostname], [chosen port])
I have an application to transfer data from remote systems to HDFS using map reduce . I however am lost when I have to deal with isues like network failure .. That is , when a connection from remote data source is lost and data is no longer accessible to my mapreduce application. I can always restart the job but when data is huge then restarting is an expensive option . I know the mapreduce would create temp folder but will it put data there ? Can I read that data out and then Can I somehow start reading the rest of the data ?
A mapreduce job can write arbitrary files, not only the ones managed by Hadoop.
Configuration conf = new Configuration();
FileSystem fs = FileSystem.get(conf);
out = fs.create(new Path(fileName));
using this code you create arbitrary files which work like normal files in the local filesystem. Then, you manage connection exceptions such that when a source is unaccessible you nicely close the file and record somewhere (e.g. in HDFS itself) that happened an interruption and at which point.
In the case of FTP, you could write just the list of file paths and folders. When a job finish to download a file, write its path on the downloaded list, and when an entire folder is downloaded write the folder path, so in case of resume you will not have to traverse a directory content to check that all files were downloaded.
At the program startup, on the other hand, it will check this file to decide whether the previous attempt failed and, in case, where to start the download.
In general, Hadoop will kill your program if it's not writing/reading anything for a timeout. Your application can tell it to wait but in general is not good to have an idle job, so it's better to end the job nicely instead that waiting for the network to work again.
You can also create your own filewriter, this way:
conf.setOutputFormat(MyOwnOutputFormat.class);
your filewriter could save its own temporary files in the format you prefer, so if the application crashes you know how files are saved.
HDFS saves files with chunks of 64MB by default, and when a job fails you may not even have a temporary file unless you use your own writer.
This is a generic solution, it depends on which is the source of data (ftp, samba, http...) and its support to download resumes.
EDIT: in case of FTP, you could just use csync to syncronize a FTP server with your local filesystem, and hdfs-fuse to mount a HDFS filesystem. It works when you have many small files.
You haven't specified what tool you are using to ingress data into HDFS/Hadoop.
Some of the tools that you can use to ingress data into HDFS/Hadoop which support recoverability are Flume, Scribe & Chukwa (for log files) and they all support various configurable levels of file transfer reliability guarantees, and Sqoop for transferring relational db data into HDFS or Hive, etc.
I need a system to analyze large log files. A friend directed me to hadoop the other day and it seems perfect for my needs. My question revolves around getting data into hadoop-
Is it possible to have the nodes on my cluster stream data as they get it into HDFS? Or would each node need to write to a local temp file and submit the temp file after it reaches a certain size? and is it possible to append to a file in HDFS while also running queries/jobs on that same file at the same time?
Fluentd log collector just released its WebHDFS plugin, which allows the users to instantly stream data into HDFS. It's really easy to install with ease of management.
Fluentd + Hadoop: Instant Big Data Collection
Of course you can import data directly from your applications. Here's a Java example to post logs against Fluentd.
Fluentd: Data Import from Java Applications
A hadoop job can run over multiple input files, so there's really no need to keep all your data as one file. You won't be able to process a file until its file handle is properly closed, however.
HDFS does not support appends (yet?)
What I do is run the map-reduce job periodically and output results to an 'processed_logs_#{timestamp}" folder.
Another job can later take these processed logs and push them to a database etc. so it can be queried on-line
I'd recommend using Flume to collect the log files from your servers into HDFS.