TLDW logo

Training - 1hr Introduction to GoldenGate 21c Microservices

By oraclegoldengate

Summary

## Key takeaways - **GoldenGate Microservices in 21.3**: GoldenGate version 21.3 covers all major builds including Oracle Database, DB2, MySQL, PostgreSQL, SQL Server, and Big Data. [02:24], [02:44] - **Four Core Microservices**: Each GoldenGate deployment runs four major services: Admin Server to manage extracts and replicats, Distribution and Receiver for trail file transfer, and Performance Metrics for monitoring. [19:19], [19:39] - **Web UI Setup via OGCA**: After Oracle Universal Installer, run OGCA script to create Service Manager, deployment, set ports, environment variables, and admin user/password via guided web interface. [34:02], [34:24] - **Distribution Path Bridges Deployments**: Source Distribution Service sends encrypted trail files via SSL/WSS to target Receiver Service, with authentication and NGINX reverse proxy exposing one secure port. [45:15], [45:35] - **Admin Client Replaces GGSci**: Admin Client offers secure remote CLI like SQL*Plus with user authentication, roles, and Linux command history, running most GGSci commands remotely. [49:52], [50:11] - **Docker Simplifies Deployment**: Build Docker image from GitHub files and installer, run container with mounted volumes for deployment and certs, accessing via mapped port like 3443. [01:03:03], [01:03:25]

Topics Covered

  • GoldenGate Born from Fault-Tolerant Tandem DNA
  • Microservices Replace Monolithic GGSci
  • WebUI Eliminates Command Memorization
  • Distribution Paths Bridge Deployments Securely
  • Docker Simplifies Cloud Deployments

Full Transcript

welcome to the oracle golden gate youtube video series oracle golden gate is the industry's premier multi-cloud technology for building real-time data fabrics data mesh and data replication solutions with class leading high

throughput and extremely low latency streaming data oracle goldengate is in a league of its own thousands of customers in over 180 countries globally trust goldengate with their most important data and goldengate

is consistently rated as a top leader for data integration and enterprise data fabrics one key reason that golden gate is one of the most successful integration tools in history is the consistent innovation

over more than 20 years goldengate has notched many many industry firsts for modern event-driven streaming data solutions golden gate is first and foremost known for its data event capabilities

including long-time leadership for data high availability transaction replication and change data capture more recently since 2015 the most successful areas of golden gate have

been in the areas of data pipelines and stream analytics modern technology solutions demand real-time data ingestion streaming etl for low latencies and best-of-breed analytics on data in motion

goldengate delivers modern time series analysis in a no code development package with enterprise class data governance including best-in-class data security data verification and rich

metadata catalog another reason that golden gate is so successful is it is truly a tool for all enterprises supporting a huge variety of

common databases applications nosql data stores clouds and messaging systems in a powerful low latency data fabric goldengate's modern microservices architecture is the foundation of

real-time data events for the event-driven business with an easy-to-use browser-based user interface and available cloud services for just pennies per hour there's nothing stopping you from realizing the future today

try goldengate for free follow this link for more information and now onwards to the video you came for hi everyone uh welcome to the presentation for the golden gate

microsoft architecture introduction my name is warner hua i am the product manager for oracle golden gate and today i'm going to share with you the

new microsoft architecture for golden gate before i go to the detail i want to share with you some news on what we have

released so far on the microservice architecture so we released golden version 21.3 and this version

covers all the major golden gate builds including golden gate for oracle database going either for db2 0s mysql

postgresql as sql server and most importantly the golden gate for big data so we are going to

be basing today's presentation on the version 21.3 to introduce the micro series architecture so here is today's agenda i'm going to

cover six sections of this presentation the first part is the brief introduction of golden gate platform

because some audience here probably are not quite familiar with golden gate so i want to give everyone a summary and overview of what goldengate is

and the next section we're going to cover why we are going to release the micro service architecture what's the benefit of that

and then the next section will be introduced the web ui for the microsoft architecture which is a important feature we put in there and it will be

liked by the new generations golden gate admins and we are also going to cover the distribution path

basically it is a way to send the golden gated data between different golden gate deployments almost like a bridges between the different golden gate

and then we are going to review the admin client address api this is the feature will be more for these old timers who are not interested in the web ui but

they are more interesting the command line interface for the golden gate the last part i'm going to cover is the docker implementation which is pretty

popular nowadays how you can build golden gate docker image how you can start a golden gate container

as a matter of fact the golden gate oci cloud implementation is actually built purely on top the golden gate docker implementation

so that's the high level overview of today's agenda so let's come to our first uh golden gate topic the

introduction for golden gate i want to give you a little bit history of how golden gate evolves to today so let's

fast backward to the 1990s that's when golden gate first come into the market uh golden gate is first developed on a platform called

tandem today this brand name probably is not very familiar with everyone as as a matter of fact tandem has been

acquired by compact and compact has been acquired by hp so in today's market the tandem is known as hp non-stop server

but back in 1990s tandem is very popular in the mission critical industry so a lot of financial industry and government agencies are using tandem to

handle their mission critical data goldengate was born in that time to help the customer to re to replicate

the data in real time on tandem so it can provide a h-a-d-r solution for the tandem customers

and as tandem system is very fort tolerant or oriented in design golden gate actually has inherited a lot

of thought tolerant programming style including the non-stop pair the checkpoint between the processes and the keyboard threads

updated and make sure there's no duplicate data there's no missing data so all these are actually started from the tandem gene has been

built in the product back in the 1990s that's why today when you look at the golden gate you still see a lot of familiar uh behaviors on the

checkpointing on the data integrity guarantee those are all coming from the old early days

um back into 2000 uh the tandem customers start to explore the capability of sending the oltp transaction to

a less expensive platform because that the ford tolerance feature of 10 really makes it's very expensive on both hardware and software

so the customer always try to look for the cost saving solutions so they

tend to replicate the data from 10 real time into a cheaper data warehouse solution uh there that is where golden gate uh

starts to roll out a new feature of sending data from the tandem uh platform to the open system we're called which is like

oracle sql server db2 site base those systems running either on windows or unix or linux and so that is the where the oltp replication feature gets added to

goldengate and then back into 2009 oracle comes to acquire the golden gate and at that time the data warehouse is a

very popular concept and the customers start to populating their data warehouse with the transactional data in real time so

goldengate started to introduce the data warehouse real-time replication concept and to make sure the data can flow

heterogeneously not only on the from tandem to other system but also between any system we support it so at that time you know oracle sql server

sidebase all become a source you can capture data from there and then send it to any platform we support it and back in 2015 uh the

cloud and the data lake uh concept have been introduced to the market and the golden gate is used to replicate the data

and send the real-time transaction data to the cloud and the data lake on this big data platform and this is also when 10 uh the

oracle started to invest a lot on golden gate to added more platform support and recently there is a new member has

been added to golden family which is the oracle streaming analytic and the golden gate acquired this tool and now we can also provide a real-time

data streaming analytic to all the customers so that is how golden gate evolves from very early days a

replication software running only on tnm to today were a heterogeneous real-time replication plus stream analytic platform

under oracle now let's take a look at all this platform golden support if you look at this chart on the left side it's all the source platform we

supported of course we start from very basic on the top left those are all the relational database like oracle mysql db2

those are the basis we uh you know start a build product back in the 1990-2000 and then we started to add a lot of the

other platform like jms like the oracle applications a majority of oracle applications are running on oracle database so

right now all these oracle applications running on a cloud you are able to using goldengate to extract the data from those database in real time and ascend

it to anywhere you want and then on the right side we can see golden gate support as a target from the

dbms to the cloud platform and all the big data which covers kafka the apaches

popular platform and also hadoop and also we added the nosql support as well like mangodb and recently we also added support for

the oracle stream analytic so goldengate can send the real-time transaction chains in the streaming

fashion into the golden extreme analytic so it can do the real-time data analysis um the most popular golden gate implementation

in the network topology wise is on the database so if you look at this chart the implementation is very straightforward it can be one-to-one

active active or one-to-many many-to-one or peer-to-peer and or you can simply use goldengate as the data distributor to

distribution the transaction data to other platforms so those are based on the rdbms and on top of that we added more

implementation model like you know data lake ingestion and the stream ingestion cloud ingestion and sas replication etc etc

and today with the stream analytics available we can actually do the real-time data transformation real-time data pipelines and the geo-fencing and

the time series analysis plus we can actually uh integrate with the predictive analytics for all the models

you uh pre-created we can import that into the streaming analytic and using the pre-built uh the pre-built model to do the data analysis

with the more than 25 years in the market golden gate is trusted by a lot of big customers

we now have more than 8 000 customers in more than 180 countries and the customers spread to multiple industry sections

from from the communications to the finance media retail and you see this chart of the logo uh you will see all these uh the

fortune uh 100 companies in there uh 90 of the global payment industry are actually using goldengate to wrap the data so they really trust goldengate because

goldengate is built to guarantee there's no data loss and also golden is the build to be running

very consistently and to deal with all the exceptions handlings plus because of all these big customers

we are investing on golden gate every day to try to beef up the performance and there's a lot of big performance numbers we achieved on oracle database

and also on another oracle we are working with the customer as well to satisfy the requirement to handle the big transaction data so you can see we can easily achieve four terabytes redo

log per hour reading speed under oracle redo log processing and we achieved a 1.5 terabyte redo log per hour at single table

and on non-oracle database we have been able to achieve four billion transaction per day on the db2 active active replication on the fortune 200 retailer

in the asia pacific we have a mega bank and we can achieve 55 000 transactions per second on memphis so those are all the numbers

showing you golden gate is really not only a reliable real-time data application platform it can also handle large transaction volume as well

so that concluded the introduction for the golden age platform and now we're going going to come into the section on why

we are trying to push the micro service architectures so talking about microservice architectures so here are the

major code chains we introduced uh the first one is improved command line interface we replaced the old ggsei with the

golden gate admin client and this atomic client is made like a sql plus client you can run the admin client at anywhere

and to remotely connect to any golden gate deployment and it's actually secured because it has its own user id and a password

authentication system so it's not like the old gsci as long as you can gain the access to the uh operating system you can gain access to golden gate not that anymore

so golden git has its own authentication system and it's locally authenticated which means for different golden gate deployments they have their

own user id and password so it's in it's independent which with each different uh golden gate deployment uh of course we introduced a very

beautiful web user interface um the this is actually a uh complaining point in the past for the older golden gate because older golden

can always only have the command interface it does not have any web ui interface and then now with the microservices we do have we do introduce the web ui interface to help

the uh new user to intuitively to uh configure to monitor goldengate

and you you can do everything all the tasks in the web ui interface um it also reduce the learning curve right so you don't have to remember all

the commands because the web ui interface basically will guide you through on each step in the same time we also enhanced the golden gate

added the rest api support so for those experienced users they can actually write the rest api and to

put that in the script to fully automate their golden gate configuration and golden in the monitoring and the web ui interface is also built to use

api so which means anything you can do on the web ui interface you can actually do that in the rest api just in the uh simple

script and this can be integrated to almost any popular platform or tools like the linux script or sql plus or

java or python talk about all these features we added to microservice architecture i want you to see how these features are actually

being built uh through all those different modules so this is how golden gate and micro services are architecturalized how they are built up

we have a service manager act as the main component to monitor all the golden gate deployments so typically on each machine you have

one going to the service manager and underneath this service manager you can have multiple golden gate deployments and each golden gate deployments will have

four major services running underneath that there's a golden gate admin server there's a golden gate distribution service golden gator matrix and the performance

service and receiver service the admin service is the most important one it actually manages all the golden gate key

component such as extract the replicat it manages all the database login alias in the credential store

the golden extrusion service and the golden gate receiver service are a pair to be used to send the golden gate trail files from one go negative deployment to

another so on your source golden gate deployment you always run the golden gate distribution service to create a distribution path and send it out to the golden gate

receiver server running in your golden gate target deployment so these two services can talk to each other and then we support the ssl encryption

so you can send the data basically from the distribution service to receiver service using the ssl the golden gate monit the monitor and

the matrix service will monitor all the key golden gate component and it shows all the threads process

status and the cache status it's shown everything in a graphic user interface to help the customer to

monitor the golden gate in real time so this is how the golden gate micro services are built up with all the different

services and then on top of that you can also run a reverse proxy service like engines the benefit of running engines is you do

not have to expose all the ports for the local services and instead engines will just open one port and then we'll handle all these other service communications

to pass the write request right service internally so this will just further protect and secure your golden gate deployment

and all this can be called and accessed from the golden gate admin client or rest api or from the web user

interface to access all the services and also for the golden gate implementation you can implement multiple golden gates or multiple

instances which we called a mesh implementation basically this is very useful when you try to

put a golden gate on multi-cloud multi-data center architecture so each cloud or each data center can have a set of golden gate

installed and between those golden gates you can have the uh trail file be passed between them uh through the ssl and so those traffic can

go not only on a dedicated connection can also go through internet as well so this is almost like a mesh network so you have golden gate on each different

node and then they can talk to each other or you can have golden gate on the uh which were called a hub configuration

so you install goldengate on one server and then you have remote connect to the source database remote connector target target database to perform the remote

capture and the remote delivery this configuration allows you to manage golden gate centralized in one place

but it requires a high network bandwidth and the low latency on the network because it uses the remote capture and the remote delivery

especially on the remote delivery part the network latency is very critical we recommend you to have anywhere

lower than five milliseconds to achieve the better performance for the remote delivery now we want to review the benefit of using the uh golden and

microservices so for the golden and micro services the most important feature is the security it is more secure and especially for the

cloud and the distributed architectures as i mentioned the golden gate introduced this distribution path concept

allow you to use ssl to encrypt the trio file while they are transferring that across network so all the communication for the golden gate

for example the distribution pass you can use wss the websocket secured service and then for the web ui

interface you can try to use https which is also ssl enabled connection to access the web ui the golden itself also have the user id

password so when you create the distribution path you can also require it provide user id and password to login to the target

so to authenticate this is the right user to connect to target before the target will receive the data and we also integrated with oracle key vault to

store the encryption keys in a separate server so you don't have to store the goldengate trill encryption key on the

particular server golden gate is running we also have improved the architecture performance upgrade and the patching mode for the golden gate deployment

because we separated the golden gate deployment out from the golden gate binaries right so that means we can

actually upgrade the golden gate binaries without affecting the deployment directory the deployment directory is always the same and then once the binary is upgraded you just need to bounce the golden gate to take

in the golden gate new binary so this will make the golden gate patching and upgrade much much easier

we also have replaced the extract pump with the distribution pass which is a multi-threaded

distribution service so this gives customer better performance when they try to send no large data and they can fully utilize the

network bandwidth we also designed for use for the mesh and hub and the remote configurations

we have a special maa design how the hot the white paper how you can configure golden gate following the oracle's maa recommendation

and also because we separated the deployment with the the binary you can actually move the deployment from one server to another

and you can run deployment also in docker in kubernetes and directly on the db server so those are all the flexibility you you have when you try to use the

golden gate micro services and also continue on the all the benefit we are

having the same matured core functions as the classic golden gate um because the extract replica component that we're using in the microservices are exactly

the same as the classic architecture we also support for multi-platforms as i announced earlier in the beginning of this slide we are pushing all the

non-oracle database build to use the micro services as well so those are the future uh of the golden

gate of the golden gate trend so which means we in the future we will only support a micro services architecture of course we do support the mis

mixed environment when they try to exchange the data between the classic architecture and the microservice architecture we do also support that make the backward uh compatible

which i'm going to cover in the next slides on how to exchange data next this will be next couple slides after that the

last slice i'm going to talk about on the micro search features is the rest api because address api is very popular nowadays in the

devops operational mode because it can allow the machine to automatically manage a certain platform certain application

with the rest api basically it's a machine managed another machine so that's why golden gate added the rest api support so you can use that to

almost go through any golden gate major function calls basically you can make all the calls from anywhere as long

as you have the network access to the golden deployment it can integrate with the oracle classware and also integrate with

oracle's acfs dbfs for disaster recovery and it supports across the board oracle non-nuclear and calling it for big data with so many

features added we are aggressively rolling out to the customer and one of the big customers is taking advantage of golden gate micro service architecture is wells fargo

and wells fargo is using golden gate to achieve a lot of business use cases for example the high availability and the real-time replication and the data synchronization

between the core operational applications and also the real-time data warehouse data feeding are all done by golden gate so in the past like

10 to 15 years wells fargo has used a lot of different golden product starting from hp non-stop to the golden gate

oracle and they even use golden gate help them to migrate off the hp non-stop and to the oracle extra data using the golden gate to achieve the

minimum downtime migration strategy so with the time going they have more and more golden gate installed in their uh entire

network so it becomes more difficult to manage all these classic architecture golden gate so after they started the golden gate microservices

they actually really like the concept of to manage the golden gate from one central point so they actually consolidated 1500

golden classic installation homes down to 250 micro services installation and totally consolidate that and then that provide them a

very nice web ui interface to monitor to configure golden gate from anywhere and also provide an address api to allow them to automate the golden

gate configurations and management so this is just a real customer example of how they are taking the advantage of golden gate microservices to

to consolidate all their golden gate installations so the next section i'm going to cover is the most important the web ui like i mentioned earlier

web ui is mainly there for the new generation because nowadays the new system admins they do not have a lot of time to

read the menu right and to memorize all the different commands have a web ui is really very important for them to have a short

learning curve and can take on the product and they can just focus on what exactly they are doing instead of trying to learn the product itself

so in the web ui installation uh it actually car used by the oracle universal installer you run the oracle

universal installer it will install the golden gate binary after that you are going to use the ogca another web ui tool to start the golden gate

configuration so this is very much like the oracle database configuration right you run the or the oracle universal installer and then you run the database configure

admin assistant to help you to configure database so here is couple of screenshots of the oracle universal installer

which been used to install goldengate and you can see it's very similar to the database installation uh user interface right so it will ask you

uh what you want to install and where you want it to be installed and then where to create the uh inventory and then at last it will give you a summary of

what has what your choices are and then you can also generate a response file if you want we can just in the future we can just run this installer

without any web ui just use the uh response file uh so this is what it looked like after it's installed then it will uh ask you

to run a uh script as the route to create all this oracle inventory but you do not have to do that you can skip that as well

if you decide to maintain your own oracle inventory so once this is installed then you have the golden gate binary files

on this particular directory you specify right so this is the where the binary is so after you get the binary in there then you have to run the og ca

which is a configuration assistant uh script and i have a small video to show you how to run that right so you change first of all you have to cd to the directory

where the golden gate the binaries are and there's a by bin direction inside the bin directory there's og ca script which you just started up it will start

the web ui interface to go through to ask you what you want to do so in our case we want to create a new service manager right you have to tell it where

the service manager's home is and to tell which port this is the tcp port the service manager is going to listen to and then it's going to check the system to make sure the port is available you know the hostname is configured

correctly and all the golden software is there then it will give you a choice to add a new golden deployment so you have specified what's the new

deployment's uh name so it will be use that name to create the deployment and keep the software home unchanged that's where the binaries are

and then you have to tell it where the deployment going to be created so after that you have to add in some environmental variables like tns admin

which you put in the sql net and the tns names files for the future database connection this can be anywhere and then you have to create the user id and

password for the admin user for the golden gate deployment so this is what i talked about earlier our own internal user authentication system to see who

can log into the golden gate system and every deployment can have their own set of the user id and password then after that it will ask you the

do you want to install a secure deployment or you do a non-secure deployment but you want to use the client have a

certificate to send the data to the another secure deployment so in this particular example we are doing the non-secured but the client will have a

particular certificate so we still can use ssl when talked to a secured deployment so once you supply all the certificate

they will continue to ask you about all the ports this deployment service services are going to be running on so once you put in the ports it will

automatically populate the other ports for all the other services and then it will ask you about the default goldengate schema name which is backward compatible

and this pretty much all you can do the finish and then we'll start doing all this work in the background the same thing you do have a choice to save the response file so next time you can just

add the response file and then it will be able to do a silent installation without the web ui so once you get all the installation done you can see the service manager is actually up and running

in this service manager and also all these services for the deployment we just created is also are up and running

so those are the four services uh admin server and distribution server uh performance monitor server and the receiver server so now you can access to

the golden gate web ui on this system because it's already up running so you go to the port 9000 that is the service managers port and then you will see you

use the goldengate user id and password the admin user id password login you will see all the goldengate services running in there and this is one

deployment we created so that pretty much covers the ogca how you use that script to configure and to start your first golden gate

deployment so once the deployment is started you can actually go to log in to the web ui interface like i just said

this is the service manager interface you logged in once you use that admin user id login you can see the deployment

and then you have to log into the admin server right that's the deployment have admin server use the same user id and password to log into the admin server

and it will give you the web ui access to the admin server to allow you to add extract at a replica and so for this one in particular we

already have extract running so you can use the web ui to check the detail of the extract and see the checkpoint information and to see the statistics how much data they capture

and also they checked the cache manager statistics how much memory used to queue up the big transactions and what you put in the parameter file

and also the report file those are all things you can then down and can be done on the web ui without logging into that golden gate server and

those are very popular tasks you can use at the same time you can check the performance metric service and also it can list out all the major component or

process um what their current status the memory usage the disk usage cpu usage all in the real time chart and you can monitor that and

also you can check on the statistics and the configurations and also the trill file how much data is written and the database

statistics cache statistics and even boundary recovery for the extract you can check on that as well and the receiver server as i mentioned earlier it is a corresponding to distribution

server so it shows you all the distribution paths we created currently are running so that basically gives you all the information you need to monitor or

configure golden gate when you access this web ui and this is where you can connect the database in the admin server once you configure the

alias to connect to the database and also here is the administrator manual led to create new users and so this is the overall and menu

which you can actually see all the critical messages or error messages uh in the this particular golden gate

deployment so this is the how the user interface uh being used and then can navigate through different components

and the most important thing is in this version 21.3 we have unified home page so from one home page you can access all these four

different services without having to going back to the service manager to click those four services one by one so right now on one

page you have all these links and the plus we added another very useful feature feature in the 21.3 for the simplified certificate

management before in old golden versions if you have to manage the certificate you have to run the uh

or or pki to manipulate the wallet which is not very user friendly and also it's very easy to make mistake if you mess up the wallet then the golden gate

will the ssl will not work anymore so right now you can use the web ui interface to view the certificate to replace the server certificate to add a multiple client certificate or add a ca

certificate and right now we actually totally disabled the wallet manipulation

so we do not recommend the user to deal with the wallet anymore and also the distribution pass can choose multiple uh client certificate if you are using the certificate

authentication method and this is what looked like on this web ui interface on the certificate management on the client certificate section you can actually

add in multiple client certificate but normally you only have one this is certificate used when you started talking to another golden gate deployment

this is the certificate you use to identify yourself and then you have the ca certificates those are all trusted certificate or the

root certificate of the other server-side certificate so once you have the trusted certificate added in here you can trust the

certificate sent by the other side which is the server side you can do the certificate delete or add or replace replacement all from this web

ui interface so this basically allows you to manage all these things from the web ui without the manipulate

directly on the uh the wallet so those are all the files where they are for the uh the golden gate like wallet report file trail files if you do

want to dig a little bit more to access all these files under the hood or how they works there's also some references i put here

and which can share with you later on all the user reference guide for golden gate micro service architecture and golden gate installation so that's pretty much concluded the web ui

interface introduction and the next section is talking about the uh trail replication we send the trail from one golden

employment to another employee golden deployment so this is a basically what i call a bridge between different golden gate deployments

so if you look at a golden gate the typical standalone golden gate you are having golden gate extract running inside the micro surface architecture you have

golden gate replicate running inside this and they are both doing the remote capture remote delivery there's no need to send any trail data out because everything is running in the same

instance right so we support oracle database 11g all the way to 21c and to access the database we only need

the sql net connection so in this is a very vanilla basic golden gate deployment which does not allow you to

does not need to send any data out but what if you have two golden gate deployment right you have going a deployment on one system and on another system have

another deployment if you want to send the data from the source goldengate to the target goldengate then you will

need to use the distribution service to create a distribution path to send the data the most popular

configuration for that is on your target system you are running a engines which is a reverse

proxy server and only expose one port to the external world so your source golden gate the distribution server is going to talk to the engines

and the communication between the engines and the distribution server server is ssl enabled but ssl is

terminated at engines and the engines pass the data to the receiver server and the receiver server will pass the data like write data in the trail file and then replica is going to pick them up

and then push them to target database so for this distribution path we're using the wss which is encrypted uh the

protocol and also we are using the target system user id and password to authenticate to make sure this distribution pass can log in to

this target to send the data and the engines as i said is just terminate the ssl and we double check that when we establish this ssl

connection we'll double check the certificate cn name the engines is using as a server certificate the cn name will match

exactly what this server name is or at least the there's a san name will match the server name so we know this is uh

when we connect to the server we know the server claims who they are we can verify that and the alias for the authentication is also needed to make sure it can connect

into the targets so those are like double safety rules we put on there before we can establish this connection start to sending data so this is a you know

golden edit trail transportation between two micro services architectures and then if you do have a classic architecture you can still send the data

to golden and micro services instead using the distribution path in your classic architecture you're basically using a remote host and a port number and

send the port put a port directly to the port where your receiver server is listening to so this will allow the classic architecture to

send the data to the microservice architecture but since the classic architecture does not support ssl we highly recommend you to encrypt the

trill file on your disk first before you push them through the network so that way they if if this network is not through the security link all the data will be

encrypted otherwise if this both architectures are sitting inside your local area network then you don't have to worry about the encryption

and the same thing if you do not have a dedicated connection you can actually use the socket proxy tunnel to facilitate

the uh the the connection from the pump extract to the receiver server this way you rely

on the ssh to encrypt your trail and then to send the trail file to the golden gate micro service architecture and this is a feature we added for the

classic architecture long time ago and then it's still working when you try to send data from classic to the micro search architecture so that concludes the

um content on the distribution path how you share the data between different golden gated deployments so the next topic will be

the admin client and the rest api as i mentioned before the admin clients rest api are really liked by the old timers or by

those golden gate system admin they really do not care about web ui but they want to keep doing things in the old way

to do the command line interface so the gtxci has been dropped from microsoft architecture but we added the admin client and both

are you know command line interface and almost all the ggsci commands can be run in the admin client

okay so this is a backward compatible we have done uh for the uh the golden gate admin so they don't have to learn a set of new commands instead if

they really knew the gsci command they can use that in the admin client but admin client actually added more secure more security features

in the the implementation it has the user authentication as i mentioned before so you can define different users and we

have different roles so have different level uh for the user the what kind of tasks they can perform it also can run remotely

to connect to any golden gate microservice architecture so you don't have to always access the system we're going to get installed it's almost like a sql plus right so you

install the admin client on any machine as long as you have tcp connection to the golden gate server you will be able to connect your golden deployment

and it supports the linux style command history and recall instead of a very cubersome you know fc command in the gsci

so all the linux system admins they can you know feel free to to explore you know to record all the commands to run those code golden commands

uh this is some example of how you run the admin client you start the admin client first and then you use the connect command to tell it where you want to connect so in

this particular example it's connecting to a local host right it's going and running locally on port 443 so this is basically

a injinx reverse proxies port and the login as the og admin and the explanation mark basically say do not check the certificate because i'm using

a self-sender certificate to set up the environment so once it connects to that it will ask you the password for the og admin and then you put a password in and then it tells you

okay so because there's only one deployment here i'm going to use this deployment as my default deployment and then you can see when you issue the info all command it's almost like the gsci

right but in addition to the extract replica we actually also shows you in this particular deployment how the four basic services are running whether they are running or they are

stopped so those are the information it gives to you it's purely command interface so here if you do not run that locally on the system you can run into any other system as long as you specify

the right hostname and the port it will be able to connect and this is a typical

example of doing an info extract detail you can see the output is exactly like the output of the gsci info extract detail command

it shows you the extract checkpoint information on the uh uh the extracts you know read the checkpoint the trail the right checkpoint and where the

report files are the checkpoint files and the parameter files and its start and the end time so those are all the information you can get from the

info detail command and also we have the rest api so when you do the rest api it can be called from anywhere again the

rule is as long as you can reach the golden gate deployment through tcp you can run any rest api

to talk to the golden deployment and it supports any rest api client you can do the curl or postman you know

the postman is very useful too you can just keep playing with that and it will test out the right syntax the right result and then you can add that into the script

and it can also embed it from a lot of other scripts like shell script java python and even the sql script we

do have the sql plus you can run the golden gate uh the rest apis um the support methods is a very popular message like other

major recipes that we support all these operations then i'm going to show you example of you know how i can run this rest api

like this is the endpoint i'm going to post and its operation is a host to basically create extract

right and the body of the rest api shows first of all the credential is where the database login user id and password how where they store the domain name and

alias so i do not need to expose the real user id and the password in this rest api i just need to refer to the alias it will pull out the user in the

password in the background and this is how i create the unit direction extract and begin now and it's the uh

the the trend log format is integrated and just fyi so starting in 21.3 we only support the integrated extract the classic extract

is deprecated so if you are still running the classic extract for example in the golden gate version 19 please make sure you upgrade to integrated extract before you

patch or upgrade to golden gate version 21 and this the next section is the registration part and detailed which

container your extract is captured from and then the targets where it is it basically targets the trail file and under the config

section is the extractor parameter file so those are all in the body which you can send in this post operation to the endpoint and the end point you can see

here is clearly clearly shows it's talking to the admin server that's why i said admin server is the very basic important service running so

all your basic golden gate tasks like create anything delete anything you know stop stop stop all is done through the

admin server so once it gets sent out you'll see it will just send the response back and then it'll tell you the extract is created so this is a

typical curl command to talk to the golden end point right so first of all you do the say cert

this is tell it because i'm using a self-signed certificate so you have to tell it no i don't just accept this self-send certificate

uh as trusted certificate and then the dash k access.config

access.config this is actually a config file which the text file have all the user id and the password in there so you don't expose your golden user id password in the

command line otherwise someone when they look for the uh history of your command will be able to see the user id password right or you can also use a uh like a save the cache

to save the user the password as well then you have to specify the endpoint for this particular one the endpoint is the mp points to monitor all the

processes so the action is a get operation and then i want the beautify that used with the python json tool so i can because the

result coming back is all in a json format right so it would be easier for us to read and this is what it looked like when it comes back so it basically shows you

all the uh performance uh the server what it pulled out from this particular golden gate deployment there's admin server

there's distribution server pm server receiver server and when they started and and also the port they are using those are all the basic information the

performance monitor service is going to tell you in the rest api so you can see this is really helpful if you want to

program this and embed this into your script to automate the golden gate management and the configuration so that concludes section about the

client interface the rest api so the last section i'm going to cover here is a docker implementation basically the docker implementation is

very popular nowadays because it's so easy to deploy that in a cloud environment so to do the uh especially for the system enemies right

so you have you can build a docker image and then you just deploy it anywhere you want and create your uh the the container uh the the docker really

pulls all the necessary files for you instead of you doing the installation from scratch if you think about old days when you install goldengate you have to install

the client library right you you first you have the operating system it should be ready install the client library and then install the golden binary and then you have to configure everything the

docker actually saved the first three steps you don't have to do any os client library and you just need to pull the build your docker image

so it has a very smaller footprint compared with the vm solution the docker container is running inside your os it's not a

highly isolated a vf instance right so you can say the docker can use up to how much memory but it will not use all this memory at once not like the

you know the virtualbox you give the memory give the cpu give the disk drive and then it will just grab all in the resource immediately so

another useful feature of the docker is you can actually map a local volume into a docker container so basically you

can have the docker itself in the image only have the binary files and then the container everything's using the local

file system file the volumes and also another advantage of the container is in with golden 21.3 it actually come with the

client already pre-installed and that client library can deal all the way from oracle 11204 up to oracle 2121c

so that's the uh basically the advantage of using docker basically just to simplify your golden gate creation and give you easy

access to a already installed configured golden gate so how to get the docker image so first of all we do not at this moment

we do not provide a doctor image what we provide is a docker build files so you can go to the oracle's github

to pull that build file from this particular link and once you put it down you also have to download the golden installation

files so with these two files ready you can basically run the docker build command to tell it uh what kind of uh

uh coding you want to build basically give the image a tag and then you have some build uh arguments you have to pass

like the particular installer where the installer file is that's the file you pull down from the hotend download side and then it will send the build contacts

to the docker demon and it will start to build the docker container using this particular download file so once the

everything's built successfully you will see it will have this image ready i have this particular tag and then you can check the image you will see

in your docker repository you will be able to see this particular docker image so at this step the image is built it's ready to run

then the next thing is you're going to run the docker container starting using this particular image so there are a lot of parameters allow you

to specify the docker run command first one is name you want to give a name to this container right and then the port because we are using the engines by

default in this docker container so you need to tell injects which port it's listening to so we normally let it listen to port 443

right because that's the uh the default uh the https port it's the nx is listening to

and then you can specify uh the golden gate admin user id and password what the user id and password is and then you have to specify the golden

deployment name and then you have to specify where you want to mount this volume the

slash u02 basically is the golden gate deployment volume so all your golden gate deployment will be mounted to this volume you can actually

map this volume to your host mount point and use slash u03 is all the temporary uh directory which you can

create it and for hosts the golden gate cache manager file golden get temporary file again the same you can mount that to your host one point as well and then the

golden gate certificate this actually this certificate use the engines so you have to create the engine certificate ahead of time and then to tell the container

hey this certificate is actually at here this small point and then this the image name of course you can also run it as a detached and all the other options

are supported as well so by doing all this the docker container will be started up and here you can see it's

actually come up and running you do a docker ps you can see the uh the docker containers on the up running here and in the dash p parameter we're mapping

the containers 443 port to the host local 3443 port so which means you can access this particular docker container of golden gate implementation using the

port 3443 so that's actually allow you to start the docker container so once you have the image you can start multiple containers

as many as you want as many as your system can handle right and once the container is up and running and this is how you attach to the container and try to run

the admin client so you first use the docker exct i t command to attach to this particular

docker container and the tell to start a bash once you start bash you're in you are in as root you have to sue to ogg user because that is where the golden gate is

actually running and then you can cd to the u01 ogg bin directory and start the admin kind and then you are inside the admin client you can connect

to the golden gate from there so once you're inside that you can connect to the localhost 443 use the golden gate admin user id and give the

password and then list out what's in this particular golden gate environment so that's just very simple steps to show you how we build a docker

image how we start a docker container from image and then how we access the container to run the golden gate utilities those are very high level steps

but you can certainly just start from there to explore all the other features using the docker container to run golden gate so that's concluded the darker

section part and that's all of my today's presentation and thanks everyone for your attention your time and if you have any questions or any comments please feel

free to share with me my name is warner and i am the product manager of the oracle golden gate thank you

[Music] you

Loading...

Loading video analysis...