Zero Downtime Migration – Migrate to Exadata DB System (ExaCS)

A while ago, Mike and I were testing upgrades on Exadata on a system that we had borrowed for a short period of time. Now, they (the owners) haven’t reclaimed the Exadata DB System yet. Some times I wonder whether they forgot all about it – and I won’t tell them because I like the idea of my very own Exadata playground. I decided to try out a migration using Zero Downtime Migration (ZDM) to an Exadata DB System (ExaCS).

You can use Zero Downtime Migration to easily migrate to an Exadata DB System (ExaCS) Exadata Cloud Service.

Prepare source environment

There are no changes to the way that you prepare your source environment. Just follow the same procedure as described in a previous blog post.

The source database can run on any Linux platform – a regular one or an Exadata. It doesn’t matter. ZDM will take care of the details for you.

Prepare target environment

To get the full benefits of Exadata you should be running RAC databases. Exadata and RAC is a perfect match but it is up to you to decide. If your source database is already a RAC database (or RAC One Node) you must migrate to a RAC database. However, if your source database is a single instance you have to option to either stay single instance or go RAC. If you go RAC you can use ZDM to make all the changes for you; simple and easy.

You must create a placeholder database on the target system. The placeholder database gets overwritten ZDM during the migration but it is initially used by ZDM to get information on how you want to configure your target database in OCI. For example, the migrated database will be placed in the same Oracle Home as the target placeholder database. Also, the architecture is determined this way. In other words, if you create the target placeholder database as a RAC database; then your source database is automatically converted to a RAC database during migration. If you create a single instance placeholder database; you get a single instance database.

Just like any other migration, when creating the placeholder database there are some things you should be aware of. On the OCI webpage you have to:

  • Set Database name to the DB_NAME of the source database.
  • Set Database version to the same as the source database.
  • Ensure the patch level of the Oracle Home match that of the source system – or be higher.
  • Ensure that the Password matches the SYS password of the source database.

When using Zero Downtime Migration (ZDM) to migrate to Exadata DB System (ExaCS) be sure to create the target placeholder database in the correct manner.

The other parameters doesn’t matter – the database gets overwritten anyway by ZDM. When using the OCI webpage to create a new database on an Exadata DB System you can’t specify exactly which Release Update to use. You will get the same patch level as the DB System Version. So you might end up having to apply a patch to the DB System – or use another tool to create the database. Also, be aware when using the OCI webpage you get a RAC database. There is no option to change it. But it is after all the perfect match for Exadata anyway.

Now, if you want more advanced options – like creating a single instance database or specifying an exact Release Update of the Oracle Home, you can’t use the OCI webpage. You will have to use either dbaaspi or dbaascli. That gives you full control over the options – but they are not as easy to use as the webpage.

How To Migrate

When your target database is a RAC database ZDM needs to connect to only one of the nodes in the cluster. It doesn’t matter which node you choose, as long as ZDM can connect to that node. And, obviously, be sure to specify that node in the targetnode parameter of the zdmcli migrate database command.

To perform the actual migration you can just follow the procedure I described in a previous blog post. That’s really all there is to it.

It’s A Wrap

I have created a video on YouTube that demos a migration to Exadata DB System.

Speaking of YouTube, I suggest that you subscribe to the Oracle Database Upgrades and Migrations YouTube channel so you never miss anything.

The Exadata Cloud Service is an awesome platform and it is really easy to migrate to it using Zero Downtime Migration. And converting to RAC is even easier.

Other Blog Posts In This Series

Zero Downtime Migration – The Pro Tips

In this final (for now, at least) blog post I will show the pro tips that might come in handy. It is a little mix and match of all my notes that didn’t make it into the previous blog posts, but are still too good to go.

Pro Tip 1: Converting To Snapshot Standby For Testing

This is a really cool feature of ZDM (or in fact any migration that uses a standby database). Once the standby database has been built in OCI and while you are waiting to do the switch-over, you can use the database in OCI for testing. So you can do realistic testing on the database you are about to switch over to. To convert the standby database to a snapshot standby database:

alter database recover managed standby database cancel;
shutdown immediate
startup mount
alter database convert to snapshot standby;
alter database open;

Now, the database is opened in READ WRITE mode and you can use it for testing. Don’t worry, the database is fully protected by flashback logs so you can always rewind any changes made, and resync with the primary database. To convert back to a physical standby database:

shutdown immediate
startup mount
alter database convert to physical standby;
shutdown immediate
startup
alter database recover managed standby database disconnect from session;

You can even do this multiple times if you want several test runs on your OCI database. If you want you can read more about the different standby databases or snapshot standby databases in particular.

Pro Tip 2: Monitoring Queries

When you have created the standby database in OCI and are waiting for the switch-over you can use these commands for monitoring. On the source/primary database:

SELECT 
   host_name, instance_name, db_unique_name, status, database_role, open_mode 
FROM
   v$database, v$instance;
SELECT thread#, max(sequence#) FROM v$archived_log GROUP BY thread#;

On the target/standby database:

SELECT 
   host_name, instance_name, db_unique_name, status, database_role, open_mode 
FROM
   v$database, v$instance;
SELECT thread#, max(sequence#) FROM v$archived_log WHERE applied='YES' GROUP BY thread#;
--MRP process should be 'APPLYING_LOG'
SELECT process, status, sequence# FROM v$managed_standby;
SELECT * FROM v$archive_gap;

Pro Tip 3: Log Files

If something goes south where can you find the log files? On the ZDM service host:

  • $ZDM_BASE/chkbase/scheduled
  • $ZDM_BASE/crsdata/[hostname]/rhp

On the source and target hosts you can also find additional log files containg all the commands that are executed by ZDM:

  • /tmp/zdm-[some number]/zdm/log

Pro Tip 4: Troubleshooting

When you a troubleshooting it is sometimes useful to get rid of all the log files and have ZDM start all over. Some of the log files get really big and are a hard to read, so I usually stop the ZDM service, delete all the log files, and restart ZDM and my troubleshooting. But only do this if there are no other jobs running than the one you are troubleshooting:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmservice stop
[zdmuser@zdm]$ rm $ZDM_BASE/crsdata/*/rhp/rhpserver.log*
[zdmuser@zdm]$ rm $ZDM_BASE/chkbase/scheduled/*
[zdmuser@zdm]$ $ZDM_HOME/bin/zdmservice start

There is also a chapter about troubleshooting in the documentation but it more or less says the same.

Pro Tip 5: Aborting A Job

Some times it is useful to completely restart a migration. If a database migration is already registered in ZDM, you are not allowed to specify another migration job. First, you have to abort the existing job, before you can enter a new migration job.

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli abort job -jobid n

Now, you can use zdmcli migrate database command again. By the way, the abort job command is missing from the CLI reference but it is a valid, and fully supported command.

Pro Tip 6: Show All Phases

A ZDM migration is split into phases, and you can have ZDM pause after each of the phases. The documentation has a list of all phases but you can also get it directly from the ZDM tool itself for a specific migration job:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli migrate database \
   -rsp ~/migrate.rsp
   ... \
   ... \
   ... \
   -listphases

Pro Tip 7: Adding Custom Scripts

You can add your own custom scripts to run before or after a phase in the migration job. You can use the -listphases command (described above) to get a list of all the phases. Then decide whether your script should run before or after that phase. This is called an action plug-in. You can bundle those together in a template to make it easier to re-use. If this is something you need, you should dive into the documentation.

Pro Tip 8: Remember To Patch On-Premises Oracle Home

If your OCI Oracle Home is running on a newer Release Update (i.e. higher patch level) then you have to patch your on-premises Oracle Home after the switch-over – and before you execute datapatch. The two patch levels should be identical after the switch-over. Release Updates are always Standby-First Installable. That means that it is allowed to have a standby database running on an Oracle Home of newer patch level – but not older one. This concept is widely used to reduce downtime during database patching and it is basically the same concept that applies for ZDM.

Pro Tip 9: Fallback To On-Premises Database

It is possible to configure ZDM to keep the on-premises database after the switch-over. It will then become a physical standby database. If something happens with the OCI database, you can do an additional switch-over and run off the original on-premises source database. This is a very nice and handy fallback method. If you want to read about it that procedure I suggest that you visit the MOS note MAA Practices for Cloud Migration Using ZDM (Doc ID 2562063.1). Be aware, falling back to the source database requires a license for Advanced Security Option. The target database in OCI is encryted using TDE Tablespace Encryption. You get that as part of any OCI DB System offering. Once the OCI database is the primary database it will generate encrypted redo – and if the source database has to apply that – it must have a license for the Advanced Security Option.

Pro Tip 10: Convert From Single Instance To RAC

A useful feature of ZDM is that it can convert a single instance database to a RAC database in OCI. And it is super simple to do that. The only thing you have to do is to create the target placeholder database as a RAC database. ZDM will detect that and take care of the rest.

Finally, let me mention that if the source database is RAC One Node or RAC, then the target database must be a RAC database. Be sure to create the target placeholder database as RAC.

Other Blog Posts In This Series

Zero Downtime Migration – Migrate Your Database

In the previous blog posts we configured the environments and installed and configured the ZDM service host. Now we can start working on the actual migration.

I will use Object Storage as the staging area between the source and target host, and, hence, I need to create a bucket that I can use for that purpose:

$ oci os bucket create \
  --compartment-id "..." \
  --name "zdm-staging"

Connect to the ZDM service host as zdmuser and have a look at a template response file. It contains a description of each of the parameters that you can use:

[zdmuser@zdm]$ more $ZDM_HOME/rhp/zdm/template/zdm_template.rsp

Now, I can create my own response file:

[zdmuser@zdm]$ vi ~/migrate.rsp

In my demo it contains the following:

#Migration method: DG_OSS - DataGuard using object storage for standby initialization
MIGRATION_METHOD=DG_OSS
#This is DB_UNIQUE_NAME of the target database, connect to the target database and execute: SELECT db_unique_name FROM v$database;
TGT_DB_UNIQUE_NAME=CDB1_fra3kw
#Name of the ASM diskgroups that I will use. To get a list of disk groups and free space, connect to target database and execute: SELECT name, free_mb, total_mb FROM v$asm_diskgroup;
TGT_DATADG=+DATA
TGT_REDODG=+RECO
TGT_RECODG=+RECO
#BACKUP_PATH should be left blank when MIGRATION_METHOD=DG_OSS
BACKUP_PATH=
#URL to OCI object storage
HOST=https://swiftobjectstorage.[region].oraclecloud.com/v1/[your-tenancy-object-storage-namespace]
#The Object Storage Bucket that will be used as a staging area
OPC_CONTAINER=zdm-staging
#In my simple demo I will skip configuration of fallback, and just shutdown the source database after the switchover
SKIP_FALLBACK=TRUE
SHUTDOWN_SRC=TRUE

If you need help figuring out what HOST should be set to, you can look in the documentation. In my demo I use the Frankfurt data center, and, thus, the region is set to eu-frankfurt-1. You can visit the API documentation for a list of regions. But often the region string is listed many places. To figure out what your tenancy object storage namespace is you can use oci os ns get. Alternatively, in the OCI web page open the Profile menu and click Tenancy: <your tenancy name>. The namespace string is listed under Object Storage Settings. To read more about have a look at the OCI documentation. Thanks to Bartlomiej Sowa for putting in a comment with this information – much appreciated!

Now, let’s start a migration evaluation. Most of the parameters are self-explanatory but you can also look in the documentation. Normally, you should leave srcauth and tgtauth to zdmauth. srcarg2 identify_file refers to the private key files that are needed to SSH to the source and target host. The backupuser is the user name that you want to use to connect to OCI object storage:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli migrate database \
  -rsp ~/migrate.rsp \
  -sourcesid CDB1 \
  -sourcenode srchost \
  -srcauth zdmauth \
  -srcarg1 user:opc \
  -srcarg2 identity_file:/home/zdmuser/.ssh/srchost \
  -srcarg3 sudo_location:/usr/bin/sudo \
  -targetnode tgthost \
  -tgtauth zdmauth \
  -tgtarg1 user:opc \
  -tgtarg2 identity_file:/home/zdmuser/.ssh/tgthost \
  -tgtarg3 sudo_location:/usr/bin/sudo \
  -targethome /u01/app/oracle/product/19.0.0.0/dbhome_1 \
  -backupuser "daniel.overby.hansen@oracle.com" \
  -eval

Shortly after you will be prompted for the SYS password to the source database and also the password for the backupuser (which is your OCI user). For the latter, please note that this password is not your user password, however, it is an auth token: When using OCI oject storage this is NOT your user password, but an auth token Also, from the output you can see the ID of your job. Use it to query the status of the job:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli query job -jobid 1

Hopefully, you will end up with a successful evaluation: When status of an evaluation is SUCCEEDED

Before we start the actual migration let me say a few words about backup during ZDM migration. You should keep regular backup strategy during migration. But avoid having ZDM backups and regular backups run at the same time. Also, if you are dealing with a RAC database be sure to put the snapshot controlfile on shared storage. Otherwise, you might get ORA-00245 errors during backups.

Back on track! Start the migration, but – VERY IMPORTANT – specify that ZDM should pause after the standby has been built and redo apply has started using -pauseafter option. If you fail to do so the switch-over will be executed as soon as the standby has been built:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli migrate database \
  -rsp ~/migrate.rsp \
  -sourcesid CDB1 -sourcenode srchost \
  -srcauth zdmauth \
  -srcarg1 user:opc \
  -srcarg2 identity_file:/home/zdmuser/.ssh/srchost \
  -srcarg3 sudo_location:/usr/bin/sudo \
  -targetnode tgthost \
  -targethome /u01/app/oracle/product/19.0.0.0/dbhome_1 \
  -backupuser "daniel.overby.hansen@oracle.com" \
  -tgtauth zdmauth \
  -tgtarg1 user:opc \
  -tgtarg2 identity_file:/home/zdmuser/.ssh/tgthost \
  -tgtarg3 sudo_location:/usr/bin/sudo \
  -pauseafter ZDM_CONFIGURE_DG_SRC

I can use the same command (zdmcli query job) to query the progress of the migration. After a while the migration will pause after ZDM_CONFIGURE_DG_SRC: Standby database is created - waiting for the final "Go"

Now the standby database has been built in OCI. Redo gets transferred from my source database to the target database in OCI and applied. If you want to verify it, you can use these queries on the source/primary database:

SELECT 
   host_name, instance_name, db_unique_name, status, database_role, open_mode 
FROM 
   v$database, v$instance;
SELECT thread#, max(sequence#) FROM v$archived_log GROUP BY thread#;

And on the target/standby database:

SELECT 
   host_name, instance_name, db_unique_name, status, database_role, open_mode 
FROM 
   v$database, v$instance;
SELECT thread#, max(sequence#) FROM v$archived_log WHERE applied='YES' GROUP BY thread#;

One of the really cool features of ZDM is that I can now use my standby database for testing in OCI – before I decide to do the switch-over. You can archieve this by converting to a snapshot standby database. I will explain this in a later blog post.

Now I can just sit back and wait. All the changes from my source environment are immediately replicated to my target database in OCI. When it is time to complete the migration and perform the switch-over, I can simply just let ZDM finalize the job. You shouldn’t worry – ZDM won’t do the switch-over until it have verified that all changes are applied on the target database.

However, to prove that it really works lets add another piece of fruit to our SALES PDB:

INSERT INTO zdmtest.items VALUES (4, 'Lemon', 4);
COMMIT;

And instruct ZDM to resume the job:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli resume job -jobid 2
[zdmuser@zdm]$ $ZDM_HOME/bin/zdmcli query job -jobid 2

In the end you will have output similar to this:

Job execution elapsed time: 3 minutes 37 seconds
ZDM_GET_SRC_INFO .............. COMPLETED
ZDM_GET_TGT_INFO .............. COMPLETED
ZDM_SETUP_SRC ................. COMPLETED
ZDM_SETUP_TGT ................. COMPLETED
ZDM_GEN_RMAN_PASSWD ........... COMPLETED
ZDM_PREUSERACTIONS ............ COMPLETED
ZDM_PREUSERACTIONS_TGT ........ COMPLETED
ZDM_VALIDATE_SRC .............. COMPLETED
ZDM_VALIDATE_TGT .............. COMPLETED
ZDM_OBC_INST_SRC .............. COMPLETED
ZDM_OBC_INST_TGT .............. COMPLETED
ZDM_BACKUP_FULL_SRC ........... COMPLETED
ZDM_BACKUP_INCREMENTAL_SRC .... COMPLETED
ZDM_DISCOVER_SRC .............. COMPLETED
ZDM_COPYFILES ................. COMPLETED
ZDM_PREPARE_TGT ............... COMPLETED
ZDM_SETUP_TDE_TGT ............. COMPLETED
ZDM_CLONE_TGT ................. COMPLETED
ZDM_FINALIZE_TGT .............. COMPLETED
ZDM_CONFIGURE_DG_SRC .......... COMPLETED
ZDM_SWITCHOVER_SRC ............ COMPLETED
ZDM_SWITCHOVER_TGT ............ COMPLETED
ZDM_SHUTDOWN_SRC .............. COMPLETED
ZDM_NONCDBTOPDB_PRECHECK ...... COMPLETED
ZDM_NONCDBTOPDB_CONVERSION .... COMPLETED
ZDM_POSTUSERACTIONS ........... COMPLETED
ZDM_POSTUSERACTIONS_TGT ....... COMPLETED
ZDM_CLEANUP_SRC ............... COMPLETED
ZDM_CLEANUP_TGT ............... COMPLETED

And now my database is migrated to the cloud. Let’s query the target database:

SELECT
   host_name, instance_name, db_unique_name, status, database_role, open_mode 
FROM 
   v$database, v$instance;

And check that all four pieces of fruit are in our SALES PDB:

SELECT * FROM zdmtest.items;

And that should be it: Welcome to OCI. Personally, I would also take a quick peek at the alert log. Just to ensure things are running smooth.

Exadata DB System

If your target environment is an Exadata DB System (or ExaCS) it is a good idea to ensure that the cloud tooling is fully up-to-date with the new database.

[root@zdm]$ dbaascli registerdb prereqs --dbname [db_name] --db_unique_name [db_unique_name]
[root@zdm]$ dbaascli registerdb begin --dbname [db_name] --db_unique_name [db_unique_name]

You can read more about it in the ZDM run book which also has a sample output from the commands.

Troubleshooting

PRCZ-4001

If you run into this error:

PRCZ-4001 : failed to execute command "/bin/uname" using the privileged execution plugin "zdmauth" on nodes "doverbyh-zdm-tgt" within 120 seconds
PRCZ-2006 : Unable to establish SSH connection to node "doverbyh-zdm-tgt" to execute command "/bin/uname"
No more authentication methods available

Check your key files. They must be in RSA/PEM format (the private key must start with -----BEGIN RSA PRIVATE KEY-----).

ZDM_OBC_INSTALL_CREDENTIALS_INVALID

If you run into this error:

srchost: 07:08:00.000: Validating object store credentials..
srchost: <ERR_FILE><Facility>PRGO</Facility><ID>ZDM_OBC_INSTALL_CREDENTIALS_INVALID</ID></ERR_FILE>

Check your credentials to OCI. Remember when prompted for the password of your OCI account it is an auth token, not your password (even though the prompt text is misleading).

ZDM_GET_SRC_INFO

If the phase ends in PRECHECK_FAILED and there is no real clue about the error, ensure that the source database host is added to the known_hosts file on the ZDM service host. Also, you can verify connectivity by trying to log on via SSH:

[zdmuser@zdm]$ ssh -i <specified-key-file> opc@<name-of-source-host>

ZDM_GET_TGT_INFO

If the phase ends in PRECHECK_FAILED and there is no real clue about the error, ensure that the target database host is added to the known_hosts file on the ZDM service host.

[zdmuser@zdm]$ ssh -i <specified-key-file> opc@<name-of-target-host>

Other Blog Posts In This Series

Zero Downtime Migration – Install And Configure ZDM

To use Zero Downtime Migration (ZDM) I must install a Zero Downtime Migration service host. It is the piece of software that will control the entire process of migrating my database into Oracle Cloud Infrastructure (OCI). The requirements are:

  • Must be running Oracle Linux 7 or newer.
  • 100 GB disk space according to the documentation. I could do with way less – basically there should be a few GBs for the binaries and then space for configuration and log files.
  • SSH access (port 22) to each of the database hosts.
  • Recommended to install it on a separate server (although technically possible to use one of the database hosts).

Create and configure server

In my example, I will install the ZDM service host on a compute instance in OCI. There are no requirements to CPU nor memory and ZDM is only acting as a coordinator – all the work is done by the database hosts – so I can use the smallest compute shape available. I am using OCI CLI* to create the compute instance which you can install on your own computer or use a Cloud Shell. But I could just as well use the web interface or REST APIs.

First, I will define a few variables that you have to change to your needs. DISPLAYNAME is the hostname of my compute instance – and also the name I see in the OCI webpage. AVAILDOM is the availability domain into which the compute instance is created. SHAPE is the compute shape:

DISPLAYNAME=zdm
AVAILDOM=OUGC:EU-FRANKFURT-1-AD-1
SHAPE=VM.Standard2.1

When I create a compute instance using the webpage these are the values: Screenshot of OCI webpage where display name, availability domain and shape are shown

In addition, I will define the OCID of my compartment, and also the OCID of the subnet that I will use. I am making sure to select a subnet that I can reach via SSH from my own computer. Last, I have the public key file:

COMPARTMENTID="..."
SUBNETID="..."
PUBKEYFILE="/path/to/key-file.pub"

Because I want to use the latest Oracle Linux image I will query for the OCID of that and store it in a variable:

IMAGEID=`oci compute image list \
   --compartment-id $COMPARTMENTID \
   --operating-system "Oracle Linux" \
   --sort-by TIMECREATED \
   --query "data[?contains(\"display-name\", 'GPU')==\\\`false\\\`].{DisplayName:\"display-name\", OCID:\"id\"} | [0]" \
   | grep OCID \
   | awk -F'[\"|\"]' '{print $4}'`

And now I can create the compute instance:

oci compute instance launch \
 --compartment-id $COMPARTMENTID \
 --display-name $DISPLAYNAME \
 --availability-domain $AVAILDOM \
 --subnet-id $SUBNETID \
 --image-id $IMAGEID \
 --shape $SHAPE \
 --ssh-authorized-keys-file $PUBKEYFILE \
 --wait-for-state RUNNING

The command will wait until the compute instance is up and running because I used the wait-for-state RUNNING option. Now, I can get the public IP address so I can connect to the instance:

VMID=`oci compute instance list \
  --compartment-id $COMPARTMENTID \
  --display-name $DISPLAYNAME \
  --lifecycle-state RUNNING \
  | grep \"id\" \
  | awk -F'[\"|\"]' '{print $4}'`
oci compute instance list-vnics \
 --instance-id $VMID \
 | grep public-ip \
 | awk -F'[\"|\"]' '{print $4}'

Prepare Host

The installation process is described in the documentation which you should visit to get the latest changes. Log on to the ZDM service host as OPC using the public IP address. By using -o ServerAliveInterval=300 I can avoid getting kicked off all the time:

ssh -o ServerAliveInterval=300 -i [key-file] opc@[ip-address]

Now, switch to root and install required packages:

[opc@zdm]$ sudo su -
[root@zdm]$ yum -y install \
  gcc \
  kernel-devel \
  kernel-headers \
  dkms \
  make \
  bzip2 \
  perl \
  glibc-devel \
  expect \
  zip \
  unzip \
  kernel-uek-devel-$(uname -r)

Create a ZDM group and user:

[root@zdm]$ groupadd zdm ; useradd -g zdm zdmuser

Make it possible to SSH to the box as zdmuser. I will just reuse the SSH keys from opc:

[root@zdm]$ cp -r /home/opc/.ssh /home/zdmuser/.ssh ; chown -R zdmuser:zdm /home/zdmuser/.ssh

Create directory for Oracle software and change permissions:

[root@zdm]$ mkdir /u01 ; chown zdmuser:zdm /u01

Edit hosts file, and ensure name resolution work to the source host (srchost) and target hosts (tgthost):

[root@zdm]$ echo -e "[ip address] srchost" >> /etc/hosts
[root@zdm]$ echo -e "[ip address] tgthost" >> /etc/hosts

Install And Configure ZDM

Now, to install ZDM I will log on as zdmuser and set the environment in my .bashrc file:

[root@zdm]$ su - zdmuser
[zdmuser@zdm]$ echo "INVENTORY_LOCATION=/u01/app/oraInventory; export INVENTORY_LOCATION" >> ~/.bashrc
[zdmuser@zdm]$ echo "ORACLE_BASE=/u01/app/oracle; export ORACLE_BASE" >> ~/.bashrc
[zdmuser@zdm]$ echo "ZDM_BASE=\$ORACLE_BASE; export ZDM_BASE" >> ~/.bashrc
[zdmuser@zdm]$ echo "ZDM_HOME=\$ZDM_BASE/zdm19; export ZDM_HOME" >> ~/.bashrc
[zdmuser@zdm]$ echo "ZDM_INSTALL_LOC=/u01/zdm19-inst; export ZDM_INSTALL_LOC" >> ~/.bashrc
[zdmuser@zdm]$ source ~/.bashrc

Create directories

[zdmuser@zdm]$ mkdir -p $ORACLE_BASE $ZDM_BASE $ZDM_HOME $ZDM_INSTALL_LOC

Next, download the ZDM software into $ZDM_INSTALL_LOC.

Once downloaded, start the installation:

[zdmuser@zdm]$ $ZDM_INSTALL_LOC/zdminstall.sh setup \
  oraclehome=$ZDM_HOME \
  oraclebase=$ZDM_BASE \
  ziploc=$ZDM_INSTALL_LOC/zdm_home.zip -zdm

And it should look something similar to this: Screenshot of a successful ZDM installation

The installation process will output some warnings which are ignorable:

[WARNING] [INS-42505] The installer has detected that the Oracle Grid Infrastructure home software at (/u01/app/oracle/zdm19) is not complete.
[WARNING] [INS-41813] OSDBA for ASM, OSOPER for ASM, and OSASM are the same OS group.
[WARNING] [INS-41875] Oracle ASM Administrator (OSASM) Group specified is same as the users primary group.
[WARNING] [INS-32022] Grid infrastructure software for a cluster installation must not be under an Oracle base directory.
[WARNING] [INS-32055] The Central Inventory is located in the Oracle base.
[WARNING] [INS-13014] Target environment does not meet some optional requirements.

And there are reference to scripts that I should execute, but this is not needed as mentioned in the documentation. ZDM piggybacks on Grid Infrastructure components which is why the messages are displayed: ZDM incorrectly informs you to run root scripts - don't run the scripts Start the ZDM service:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmservice start

Which should produce something like this: Sample output when starting ZDM service (jwcctl debug jwc) And, optionally, I can verify the status of the ZDM service:

[zdmuser@zdm]$ $ZDM_HOME/bin/zdmservice status

The ZDM service is running

Configure Network Connectivity

The ZDM service host must communicate with the source and target hosts via SSH. For that purpose I need private key files to each of the hosts. The private key files must be without a passphrase, in RSA/PEM format and I have to put them at /home/zdmuser/.ssh/[host name]. In my demo, the files are to be named:

  • /home/zdmuser/.ssh/srchost
  • /home/zdmuser/.ssh/tgthost

Ensure that only zdmuser can read them:

[zdmuser@zdm]$ chmod 400 /home/zdmuser/.ssh/srchost
[zdmuser@zdm]$ chmod 400 /home/zdmuser/.ssh/tgthost

Now, I will verify the connection. In my example I will connect to opc on both database hosts, but you can change it if you like:

[zdmuser@zdm]$ ssh -i /home/zdmuser/.ssh/srchost opc@srchost
[zdmuser@zdm]$ ssh -i /home/zdmuser/.ssh/tgthost opc@tgthost

If you get an error when connecting ensure the following:

  • The public key is added to /home/opc/.ssh/authorized_keys on both database hosts (change opc if you are connecting as another user)
  • The key files are in RSA/PEM format (the private key file should start with -----BEGIN RSA PRIVATE KEY-----)
  • The key files are without a passphrase

That’s It

Now, I have a working ZDM service host. Previously, I have prepared my source and target environments which means that I am ready to start the migration process. Stay tuned!

References

* I found this blog post by MichaƂ very usefull when figuring out how to use OCI CLI to create a compute instance.

Other Blog Posts In This Series

Zero Downtime Migration – Preparations

In the following posts I will demo how to migrate a database into OCI. To make the demo as simple as possible my source database will already be running in OCI and my target will be a VM DB System.

Prepare source environment

I will create a VM DB System that can act as my source database. I will call it srchost and the database is called CDB1. Using OCI CLI it can be done like this:

oci db system launch \
  --compartment-id "..." \
  --availability-domain "..." \
  --subnet-id "..." \
  --shape "VM.Standard2.4" \
  --cpu-core-count 4 \
  --database-edition "ENTERPRISE_EDITION" \
  --admin-password "..." \
  --ssh-authorized-keys-file "/path/to/key-file.pub" \
  --license-model "BRING_YOUR_OWN_LICENSE" \
  --db-name "CDB1" \
  --pdb-name "SALES" \
  --storage-management "ASM" \
  --node-count 1 \
  --initial-data-storage-size-in-gb 2048 \
  --display-name "SRCHOST" \
  --hostname "SRCHOST" \
  --db-version "19.0.0.0"

Once completed, I can log on as opc and switch to root and copy the authorized_keys file to allow SSH connection as oracle:

[opc@srchost]$ sudo su -
[root@srchost]$ rm -f /home/oracle/.ssh/authorized_keys
[root@srchost]$ cp /home/opc/.ssh/authorized_keys /home/oracle/.ssh/authorized_keys
[root@srchost]$ chown -R oracle:oinstall /home/oracle/.ssh

Now, I will ensure that the database is in ARCHIVELOG mode; it is a requirement for using Data Guard:

SELECT log_mode FROM v$database;

Since my database is running 12.2 or higher I have to ensure that there is a TDE Keystore. This applies even if the database is not encrypted. You don’t need a license to create a TDE Keystore – not until you start encrypting stuff. The keystore must be OPEN and the type is either AUTOLOGIN, LOCAL_AUTOLOGIN or PASSWORD. In a CDB, this applies to CDB$ROOT and all PDBs:

SELECT con_id, status, wallet_type FROM v$encryption_wallet;

If status is OPEN_NO_MASTER_KEY it means that no TDE master encryption key has been created and you will need to create one. You can also find instructions on how to create a TDE keystore in the ZDM documentation.

Prepare target environment

In my demo I will be targeting a VM DB System which I will call tgthost. For the migration to work, I have to create a placeholder database on the target host. During migration the placeholder database is overwritten by the source database. On a VM DB System this is easy; I just use the database that gets created automatically as placeholder.

DB_NAME must be the same in the two databases, and hence in my demo I must set db-name in the OCI CLI command below to CDB1 (the name of my source database). Contrary, DB_UNIQUE_NAME must be different but this will likely not be a problem because OCI automatically generates a semi-unique DB_UNIQUE_NAME (if you are a strong believer in Murphy’s Law you should double-check it).

The target database patch level can be the same or higher than the source. If higher, you just have to manually execute datapatch after the switch-over (I will put this information into a later blog post). By setting db-version to 19.0.0.0 I automatically get the latest available Release Update.

Also, the SYS password has to match in the two databases, so be sure to set admin-password to your source database SYS password. Finally, you should set storage-management to ASM. This will cause OCI to install Grid Infrastructure as well and set up a SCAN listener.

In my demo I end up with this OCI CLI command:

oci db system launch \
  --compartment-id  "..." \
  --availability-domain "..." \
  --subnet-id "..." \
  --shape "VM.Standard2.4" \
  --cpu-core-count 4 \
  --database-edition "ENTERPRISE_EDITION" \
  --admin-password "..." \
  --ssh-authorized-keys-file /path/to/key-file.pub \
  --license-model "BRING_YOUR_OWN_LICENSE" \
  --db-name "CDB1" \
  --pdb-name "SALES" \
  --storage-management "ASM" \
  --node-count 1 \
  --initial-data-storage-size-in-gb 2048 \
  --display-name "tgthost" \
  --hostname "tgthost" \
  --db-version "19.0.0.0"

As opc, switch to root and enable oracle to log on via SSH:

[opc@tgthost]$ sudo su -
[root@tgthost]$ rm -f /home/oracle/.ssh/authorized_keys
[root@tgthost]$ cp /home/opc/.ssh/authorized_keys /home/oracle/.ssh/authorized_keys
[root@tgthost]$ chown -R oracle:oinstall /home/oracle/.ssh

The requirements to the TDE Keystore applies to the target environment as well, but since I am using an OCI DB System that will be setup up correctly when the system is created.

Configure Connectivity

I need to ensure that the source host can resolve the network name of the target host. It is important to add two entries – one with the host name and one with the SCAN name (they should both point to the target host):

[root@srchost]$ echo -e "[ip address] tgthost" >> /etc/hosts
[root@srchost]$ echo -e "[ip address] tgthost-scan" >> /etc/hosts

Also, I need to ensure that I can connect to the target database from the source host over SQL*Net. Just use the default service of the target CDB, and you can get it using lsnrctl:

[oracle@tgthost]$ lsnrctl status
[oracle@srchost]$ sqlplus system@tgthost-scan/[target-cdb-service-name]

Now, I do it the other way around. Ensure that the target host can resolve the network name of the source host:

[root@tgthost]$ echo -e "[ip address] srchost" >> /etc/hosts
[root@tgthost]$ echo -e "[ip address] srchost-scan" >> /etc/hosts

Ensure that I can connect to the source database from the target host:

[oracle@srchost]$ lsnrctl status
[oracle@tgthost]$ sqlplus system@srchost-scan/[source-cdb-service-name]

Create Sample Data

I will connect to the source database:

[oracle@srchost]$ export ORACLE_SID=CDB1
[oracle@srchost]$ export ORACLE_PDB_SID=SALES
[oracle@srchost]$ sqlplus / as sysdba

Note, that ORACLE_PDB_SID works from 18.5.0 and beyond. For lower versions you can just use ALTER SESSION SET CONTAINER=SALES instead. And create some sample data, so I can verify the migration actually works:

CREATE USER zdmtest IDENTIFIED BY [your-password];
GRANT CREATE TABLE, CONNECT TO zdmtest;
GRANT SELECT ON v_$instance TO zdmtest;
GRANT SELECT ON v_$database TO zdmtest;
ALTER USER zdmtest DEFAULT TABLESPACE users;
ALTER USER zdmtest QUOTA UNLIMITED ON users;

CREATE TABLE zdmtest.items (
   id      NUMBER,
   item    VARCHAR2(20),
   price   NUMBER
);
INSERT INTO zdmtest.items VALUES (1, 'Apple', 2);
INSERT INTO zdmtest.items VALUES (2, 'Banana', 1);
INSERT INTO zdmtest.items VALUES (3, 'Orange', 4);
COMMIT;

That’s It

My source database is now ready to be migrated to OCI. It will be migrated to a new VM DB System that I just created. In the next blog post I will install and configure ZDM. Stay tuned!

Other Blog Posts In This Series

Zero Downtime Migration

When you need to migrate into OCI we have a cool – and free – tool that you can use: Zero Downtime Migration (ZDM). In short: ZDM builds a standby database in OCI and when you are ready to migrate, all it takes is a simple switch-over. My fellow Product Manager, Ricardo Gonzalez, has made a good video that gives you a perfect introduction in less than two minutes.

In this blog post series I will take you through the entire process using version 19.2 (the latest at time of writing). In the end you will know all there is to know – and you can start migrating your databases into OCI. Granted, in this first blog post there will be a lot of text but I want you to have all the details. Take a deep breath – here we go!

You can use Zero Downtime Migration to easily migrate to Oracle Cloud Infrastructure - OCI

Platforms

Your source database can be located on-premises, in OCI Classic (you know, our old cloud) or already in OCI. The latter scenario is useful when you want to migrate between regions or locations or between system types (like from Virtual Machine to Bare Metal).

The operating system must be Linux. No other platforms are supported. And you can migrate into one of these platforms:

  • Virtual Machine DB System
  • Bare Metal DB System
  • Exadata DB System
  • Exadata Cloud at Customer

Release And Edition

You can use ZDM if the database is 11.2.0.4 or newer and the target release will be the same. If you need to migrate into a newer database release, you must manually upgrade the database afterwards, or use another approach. However, it is possible to migrate to a higher patch level. You just have to manually execute datapatch afterwards.

If your source database is Enterprise Edition the database will be migrated with zero downtime. For Standard Edition the source database will be offline during the entire migration. The reason is that Data Guard is used behind the scenes, and it is not licensed on Standard Edition.

Last, you can’t migrate between editions, e.g. from Standard Edition to Enterprise Edition.

Architecture

Your source database can be non-CDB or CDB.

  • Non-CDB databases will be migrated to non-CDB. In a future release of ZDM we plan to include PDB conversion as well.
  • CDBs will be migrated as-is -meaning with all PDBs.

Your source database can be single instance, RAC One Node or RAC.

  • Single instance databases are migrated to single instance or, optionally, to a RAC database.
  • RAC One Node and RAC databases are always migrated into a RAC database.

Encryption

Your source database can be encrypted, but it is not required. Unencrypted databases will be encrypted on-the-fly once they are created in OCI. The initial backup of the source database is sent over an encrypted connection to OCI Object Storage, and redo are transferred over encrypted SQL*Net.

Be aware, for unencrypted databases (or in fact databases that are not licensed for Advanced Security Option) you are not allowed to fallback to the source environment after the switch-over. The target environment will be encrypted using TDE Tablespace Encryption, and, thus, the redo will be encrypted. To decrypt the redo at the source environment – you need a license for Advanced Security Option.

Network Connectivity

  • The ZDM service host needs SSH access (22) to both the source and target database host.
  • The source and target database host needs access to OCI Object Storage over HTTPS (443).
  • SQL*Net connection (1521) are needed between the two databases hosts. In fact, you only need connection from the target database host and back to the source, if you want the option of being able to fallback to the source database host after the switch-over.

References

In case you want to read more here are some useful links:

Other Blog Posts In This Series

This is the introduction blog post in this series. Over the next days the other blog posts will follow. Stay tuned!