My team is planning for a DR solution and we need to sync data between Greenplum Databases in Production and DR sites.
We are running the 6.4 community edition. So tools like gpbackup and gprestore are not available.
pg_dump and pg_restore not an option because there is large data set involved. What is most suitable solution for our scenario?
gpbackup and gprestore is one way Greenplum users commonly keep two clusters in sync.
While gpbackup and gprestore doesn't ship with open source Greenplum Database, the tools are open source themselves and freely available from their own repository: https://github.com/greenplum-db/gpbackup
Due to Greenplum's distribution of data across segments, there is a requirement the DR cluster contain the same # of primary segments for a successful restore (although the # of segment hosts could differ).
A common approach we see Greenplum users implementing is backing up off cluster to a third party storage system (NFS, s3 compatible storage, etc..) and restoring to the destination/DR cluster from there.
There is an open source gpbackup_s3_plugin available here: https://github.com/greenplum-db/gpbackup-s3-plugin
Let us know if you have any other questions.
oak
Related
We are going to move from SQL server to Snowflake as our target database for the warehouse.
Today we have most of our ETL development done in ODI (Oracle Data Ingegrator).
So I'm intressted in to know if anyone is using ODI together with Snowflake and how it's woking.
And what experince/recommendations you have of other ETL tools together with Snowflake as target.
For example
Matillion
DBT
Xplenty
Today we have started with using NIFI moving the data from source to Azure blob storage.
But we are not sure if ODI is the right tool for the rest when we are in the cloud.
I'm really looking forward to see all your answers
Snowflake supports both transformations during (ETL) or after loading (ELT).
Snowflake works with a wide range of data integration tools, including Informatica, Talend, Tableau, Matillion and others.
In data engineering, new tools and self-service pipelines are eliminating traditional tasks such as manual ETL coding and data cleaning companies. With easy ETL or ELT options via Snowflake, data engineers can instead spend more time working on critical data strategy and pipeline optimization projects.
With a Snowflake as your data lake and data warehouse, ETL can be effectively eliminated, as no pre-transformations or pre-schemas are needed.
In addition, Snowflake Snowpark is designed to make building complex data pipelines a breeze and to allow developers to interact with Snowflake directly without moving data. Read more about Snowpark here.
https://www.snowflake.com/trending/etl-tools
If you started to transfer data from the source to Azure blob storage, I assume that you have a subscription in Azure and it is possible that Snowflake itself is placed in the Azure region.
In this case, I recommend using Azure Data Factory directly, so you have everything on one provider and support for data migration from SQL Server.
Link to documentation: Copy and transform data in Snowflake using Azure Data Factory
I am using GreenPlum 6.x and facing issues while performing backup and recovery. Does we have any tool to take the physical backup of whole cluster like pgbackrest for Postgres, further how can we purge the WAL of master and each segment as we can't take the pg_basebackup of whole cluster.
Are you using open source Greenplum 6 or a paid version? If paid, you can download the gpbackup/gprestore parallel backup utility (separate from the database software itself) which will back up the whole cluster with a wide variety of options. If using open source, your options are pretty much limited to pgdump/pgdumpall.
There is no way to purge the WAL logs that I am aware of. In Greenplum 6, the WAL logs are used to keep all the individual postgres engines in sync throughout the cluster. You would not want to purge these individually.
Jim McCann
VMware Tanzu Data Engineer
I would like to better understand the the issues you are facing when you are performing your backup and recovery.
For Open Source user of the Greenplum Database, the gpbackup/gprestore utilities can be downloaded from the Releases page on the Github repo:
https://github.com/greenplum-db/gpbackup/releases
v1.19.0 is the latest.
There currently isn't a pg_basebackup / WAL based backup/restore solution for Greenplum Database 6.x
WAL logs are periodically purged (as they get replicated to mirror and flushed) from master and segments individually. So, no manual purging is required. Have you looked into why the WAL logs are not getting purged? One of the reasons could be mirrors in cluster is down. If that happens WAL will continue mounting on primary and won't get purged. Perform select * from pg_replication_slots; for master or segment for which WAL is building to know more.
If the cause for WAL build is due replication slot as for some reason is mirror down, can use guc max_slot_wal_keep_size to configure max size WAL's should consume, after that replication slot will be disabled and not consume more disk space for WAL.
I am looking for a way to stream entire log files from edge nodes to Hadoop. To sum up the use case:
We have applications that produce log files ranging from a few MB to hundreds of MB per file.
We do not want to stream all the log events as they occur.
Pushing the log files in their entirety after they have written completely is what we are looking for (written completely = got moved into another folder for example... this is not a problem for us).
This should be handled by some kind of lightweight agents on the edge nodes to the HDFS directly or - if necessary - an intermediate "sink" that will push the data to HDFS afterwards.
Centralized Pipeline Management (= configuring all edge nodes in a centralized manner) would be great
I came up with the following evaluation:
Elastic's Logstash and FileBeats
Centralized pipeline management for edge nodes is available, e.g. one centralized configuration for all edge nodes (requires a license)
Configuration is easy, WebHDFS output sink exists for Logstash (using FileBeats would require an intermediate solution with FileBeats + Logstash that outputs to WebHDFS)
Both tools are proven to be stable in production-level environments
Both tools are made for tailing logs and streaming these single events as they occur rather than ingesting a complete file
Apache NiFi w/ MiNiFi
The use case of collecting logs and sending the entire file to another location with a broad number of edge nodes that all run the same "jobs" looks predestined for NiFi and MiNiFi
MiNiFi running on the edge node is lightweight (Logstash on the other hand is not so lightweight)
Logs can be streamed from MiNiFi agents to a NiFi cluster and then ingested into HDFS
Centralized pipeline management within the NiFi UI
writing to a HDFS sink is available out-of-the-box
Community looks active, development is lead by Hortonworks (?)
We have made good experiences with NiFi in the past
Apache Flume
writing to a HDFS sink is available out-of-the-box
Looks like Flume is more of a event-based solution rather than a solution for streaming entire log files
No centralized pipeline management?
Apache Gobblin
writing to a HDFS sink is available out-of-the-box
No centralized pipeline management?
No lightweight edge node "agents"?
Fluentd
Maybe another tool to look at? Looking for your comments on this one...
I'd love to get some comments about which of the options to choose. The NiFi/MiNiFi option looks the most promising to me - and is free to use as well.
Have I forgotten any broadly used tool that is able to solve this use case?
I experience similar pain when choosing open source big data solutions, simply that there are so many paths to Rome. Though "asking for technology recommendations is off topic for Stackoverflow", I still want to share my opinions.
I assume you already have a hadoop cluster to land the log files. If you are using an enterprise ready distribution e.g. HDP distribution, stay with their selection of data ingestion solution. This approach always save you lots of efforts in installation, setup centrol managment and monitoring, implement security and system integration when there is a new release.
You didn't mention how you would like to use the log files once they lands in HDFS. I assume you just want to make an exact copy, i.e. data cleansing or data trasformation to a normalized format is NOT required in data ingestion. Now I wonder why you didn't mention the simplest approach, use a scheduled hdfs commands to put log files into hdfs from edge node?
Now I can share one production setup I was involved. In this production setup, log files are pushed to or pulled by a commercial mediation system that makes data cleansing, normalization, enrich etc. Data volume is above 100 billion log records every day. There is an 6 edge nodes setup behind a load balancer. Logs are firstly land on one of the edge nodes, then hdfs command put to HDFS. Flume was used initially but replaced by this approach due to performance issue.(it can very likely be that engineer was lack of experience in optimizing Flume). Worth to mention though, the mediation system has a managment UI for scheduling ingestion script. In your case, I would start with cron job for PoC then use e.g. Airflow.
Hope it helps! And would be glad to know your final choice and your implementation.
I have a option of using Sqoop or Informatica Big Data edition to source data into HDFS. The source systems are Tearadata, Oracle.
I would like to know which one is better and any reason behind the same.
Note:
My current utility is able to pull data using sqoop into HDFS , Create Hive staging table and archive external table.
Informatica is the ETL tool used in the organization.
Regards
Sanjeeb
Sqoop
Sqoop is capable of performing full and incremental loading from Oracle/Teradata.
Sqoop does parallel copy of data from source systems.
Sqoop scripts can be custom genrated and scheduled by Oozie.
Open source solution for any size cluster. No license cost.
Informatica
Best Interface in ETL Industry to manage mappings.
Does not provide parallel copy options. Provides Hive mode for parallel processing. Basically converts transformation into Hive queries for execution. Also supports push downs to generate MR code.
Licensing cost per node. If you plan 500 Hadoop nodes for future data storage you need to pay 10 times as compared with 50 node cluster when you scale cluster.
Informatica BDE is relatively new product in market. INFA Developer will be usefull for working on Big data. There are challenges in supporting all latest Hadoop platform features on Informatica, also traditional RDBMS features like Sequence generation, Stateful mapping,Sessions, Lookup Transformation in Informatica BDE.
Informatica MDM does not support Hadoop.
If price is criteria for decision making, go for Sqoop. If you want to leverage flexibility of switching Hadoop plaftorm tools, use Sqoop(Sqoop project is also thinking of moving over Spark).
If you are tied to Informatica for some reason, go for Informatica. But most Informatica developers want to move to Hadoop technologies.
Although this was asked an year ago, sharing new features in Informatica
Informatica BDM version 10.1 supports Sqoop connectivity i.e. you can use Sqoop to read the data from RDBMS and load it into Hadoop/Hive
Also, there are many new features in BDM version 10.2, especially the parameterization support in the developer tool and dynamic mappings.
Tool versus handcoding was always there.
Informatica tool gives enterprise level solution which is easier to maintain.
BDM 10.1.1 supports sqoop with spark engine. Spark 2.0.1 is supported in this version so performance its pretty good.
BDM 10.2 is just released with new features like stateful variable support which was missing in earlier versions.
SQOOP must be used for the Data exchange. You have lot of options with which you can have an optimal performance. Also if you are trying to exchange the data between RDBMS(Teradata / Oracle) <-> Informatica <-> Hadoop cluster then the data would first need to be brought to the Informatica Server which may involve additional I/O.
If the data processing must be done within hive Informatica BDE must be used.
I have a requirement to ingest the data from an Oracle database to Hadoop in real-time.
What's the best way to achieve this on Hadoop?
The important problem here is getting the data out of the Oracle DB in real time. This is usually called Change Data Capture, or CDC. The complete solution depends on how you do this part.
Other things that matter for this answer are:
What is the target for the data and what are you going to do with it?
just store plain HDFS files and access for adhoc queries with something like Impala?
store in HBase for use in other apps?
use in a CEP solution like Storm?
...
What tools is your team familiar with
Do you prefer the DIY approach, gluing together existing open-source tools and writing code for the missing parts?
or do you prefer a Data integration tool like Informatica?
Coming back to CDC, there are three different approaches to it:
Easy: if you don't need true real-time and have a way to identify new data with an SQL query that executes fast enough for the required data latency. Then you can run this query over and over and ingest its results (the exact method depends on the target, the size of each chunk, and the preferred tools)
Complicated: Roll your own CDC solution: download the database logs, parse them into series of inserts/updates/deletes, ingest these to Hadoop.
Expensive: buy a CDC solution, that does this for you (like GoldenGate or Attunity)
Expanding a bit on what #Nickolay mentioned, there are a few options, but the best would be too opinion based to state.
Tungsten (open source)
Tungsten Replicator is an open source replication engine supporting a variety of different extractor and applier modules. Data can be extracted from MySQL, Oracle and Amazon RDS, and applied to transactional stores, including MySQL, Oracle, and Amazon RDS; NoSQL stores such as MongoDB, and datawarehouse stores such as Vertica, Hadoop, and Amazon rDS.
Oracle GoldenGate
Oracle GoldenGate is a comprehensive software package for real-time data integration and replication in heterogeneous IT environments. The product set enables high availability solutions, real-time data integration, transactional change data capture, data replication, transformations, and verification between operational and analytical enterprise systems. It provides a handler for HDFS.
Dell Shareplex
SharePlex™ Connector for Hadoop® loads and continuously replicates changes from an Oracle® database to a Hadoop® cluster. This gives you all the benefits of maintaining a real-time or near real-time copy of source tables
Apache Sqoop is a data transfer tool to transfer bulk data from any RDBMS with JDBC connectivity(supports Oracle also) to hadoop HDFS.