Press "Enter" to skip to content

Category: Hadoop

Finding the Right Disk and Data Node Sizes in HDFS

Published 2020-02-11 by Kevin Feasel

Lokesh Jain has some advice when it comes to disk and data node size:

There are two factors to keep in mind when choosing node capacity. These will be discussed in detail in the next sections.

1. Large Disks – total node capacity being the same, using more disks is better as it yields higher aggregate IO bandwidth.
2. Dense Nodes – as nodes get denser, recovery after node failure takes longer.

These factors are not HDFS-specific and will impact any distributed storage service that replicates data for redundancy and serves live workloads.

Click through for specific advice on maximum disk and node sizes.

Comments closed

Unit Testing in Apache Flink

Published 2020-02-11 by Kevin Feasel

Kartik Khare has a guide to help us write unit tests for our Apache Flink code:

Writing tests for process functions, that work with time, is quite similar to writing tests for stateful functions because you can also use test harness. However, you need to take care of another aspect, which is providing timestamps for events and controlling the current time of the application. By setting the current (processing or event) time, you can trigger registered timers, which will call the onTimer method of the function

Click through for demos and more details on the test harness.

Comments closed

Apache Flink 1.9.2 Released

Published 2020-02-10 by Kevin Feasel

Hequn Cheng announces Apache Flink 1.9.2:

The Apache Flink community released the second bugfix version of the Apache Flink 1.9 series.

This release includes 117 fixes and minor improvements for Flink 1.9.1. The list below includes a detailed list of all fixes and improvements.

We highly recommend all users to upgrade to Flink 1.9.2.

Read on to see all of the things they’ve fixed and improved.

Comments closed

Delta Lake and ACID Properties

Published 2020-02-06 by Kevin Feasel

Kundan Kumarr notes that Spark’s Delta Lake allows for ACID transactions:

DeltaLog is the crux of Delta Lake which ensures atomicity, consistency, isolation, and durability of user-initiated transactions. DeltaLog is an ordered record of transactions. Every transaction performed since the inception of Delta Lake Table, has an entry in the DeltaLog (also known as the Delta Lake transaction log). It acts as a single source of truth, giving users access to the last version of a DeltaTable’s state. It provides serializability, the strongest level of isolation level. Let’s see how DeltaLog ensures ACID Transactions.

Click through for the explanation.

Comments closed

Quick Hits on Azure Databricks Performance

Published 2020-02-05 by Kevin Feasel

Rayis Imayev has a few thoughts on optimizing delta table-based workloads in Azure Databricks:

2) Enable the Delta cache – spark.databricks.io.cache.enabledtrue
There is a very good resource available on configuring this Spark config setting: https://docs.microsoft.com/en-us/azure/databricks/delta/optimizations/delta-cache

And this will be very helpful in your Databricks notebook’s queries when you try to access a similar dataset multiple times. Once you read this dataset for the first time, Spark places it into internal local storage cache and will speed up the process of further referencing it for you.

Click through for several more along these lines.

Comments closed

Cloudera and R

Published 2020-01-30 by Kevin Feasel

Ian Cook shows us how Cloudera works for R users:

Our customers love it when they can use familiar syntax to work with data regardless of its size or its source. The popularity of sparklyr is a case in point: it enables R users to use either SQL or dplyr—both familiar to most R users—to work with large-scale data using Apache Spark. Two R packages developed at Cloudera—implyr and tidyquery—aim to provide this same choice of either SQL or dplyr when querying tables with Apache Impala and when manipulating R data frames.

The implyr package is new to me, but looks interesting.

Comments closed

What’s New with Delta Lake

Published 2020-01-30 by Kevin Feasel

Denny Lee and Tathagata Das announce Delta Lake 0.5.0:

With the following pull requests, you can now run even more Delta Lake operations concurrently. With finer grain conflict detection, these updates make it easier to run complex workflows on Delta tables such as:

– Running deletes (e.g. for GDPR compliance) concurrently on older partitions while newer partitions are being appended.
– Running file compactions concurrently with appends.
– Running updates and merges concurrently on disjoint sets of partitions.

Click through for the full changelog.

Comments closed

Spark is Not ACID Compliant

Published 2020-01-27 by Kevin Feasel

Kundan Kumarr explains how it is that Apache Spark is not ACID compliant:

Atomicity states that it should either write full data or nothing to the data source when using spark data frame writer. Consistency, on the other hand, ensures that the data is always in a valid state.

As evident from the spark documentation below, it is clear that while saving data frame to a data source, existing data will be deleted before writing the new data. But in case of job failure, the original data will be lost or corrupted and no new data will be written.

Click through for an explanation of these two along with a demo, and then an explanation of how Spark Datasets don’t follow the Isolation or Durability properties either. I don’t think any of this is earth-shattering to people, but it is a good reminder that Spark doesn’t fit all use cases.

Comments closed

Updates in Confluent Platform 5.4

Published 2020-01-27 by Kevin Feasel

Tim Berglund takes us through what has changed in Confluent Platform 5.4:

Role-Based Access Control (RBAC)

Back in July, we announced the preview for RBAC as part of the Confluent Platform 5.3 release. After gathering feedback and learning from everyone who tried it out, we are now pleased to announce the availability of RBAC in Confluent Platform 5.4. You can now make use of this feature in production environments with Confluent’s full support.

RBAC offers a centralized security implementation for enabling access to resources across the entire Confluent Platform with just the right level of granularity. You can control permissions you grant to users and groups to specific platform resources, starting at the cluster level and moving all the way down to individual topics, consumers groups, or even individual connectors. You do this by assigning users or groups to roles. This gets you out of the game of managing the individual permissions of a huge number of principals—a real problem for large enterprise deployments.

RBAC delivers comprehensive authorization enforced via all user interfaces (Confluent Control Center UI, CLI, and APIs), and across all Confluent Platform components (Control Center, Schema Registry, REST Proxy, MQTT Proxy, Kafka Connect, and KSQL). Given the distributed architecture not only of Apache Kafka but also of other platform components like Connect and KSQL, having a single framework to centrally manage and enforce security authorizations across all the components is, in a word, essential for managing security at scale.

Click through for several more features and where you can try it out, either on-premises or in a major cloud host.

Comments closed