Dontcheff

Cloud Nine

In Cloud, DBA, IaaS, Oracle database on May 3, 2017 at 16:58

“Get happiness out of your work or you may never know what happiness is.” — Elbert Hubbard

According to Amazon and quoted by Fortune Magazine in a recent article entitled “Amazon Data Center Chief schools Oracle CEO on Cloud claims“, AWS executive Andy Jassy said (at AWS Re:Invent two years ago) that every database customer he talks to is unhappy with their vendor: “I haven’t met a database customer that is not looking to flee their vendor.”

Another interesting article by James Hamilton entitled “How many Data Centers needed world-wide” discusses more or less the same topic. Reading the comments after it is worthwhile. Also consider reading these extensive performance results about Cloud performance and TCO of the Oracle database.

A young and extremely smart analyst from my company asked me last week: “Why is the Oracle database better than MySQL or MongoDB?”. Tough question, right? You may ask the same about DB2 or SQL Server. All databases have their pros and cons. And we as people have our preferences, based on experience, knowledge and prejudices.

If you try to find out the explanation of the quote statement on top, you might very like end up with this one: “You have to spend most of your life working, so if you’re unhappy at your work you’re likely to always be unhappy”.

So, I have been happy (if that is the right word) working with the Oracle database. Unlike DB2, you have all the tools, options and automation to tune it. With about couple of hundred MySQL databases at Nokia, we spent more time (thank you Google!) investigating issues than with more than one thousand Oracle databases. SQL Server: if you prefer using the mouse instead of the keyboard, then this is the right database for you! Teradata compared to Exadata: let me not start…

As Forrester say, In-Memory Databases are driving next-generation workloads and use cases. Check out this recent comparison of all vendors.

But back to Cloud. Have a look at what speed and what features Oracle is embedding into its Cloud. By far the best Cloud for Oracle workloads! All these are new additions to the Oracle IaaS:

What’s New for Oracle Compute Cloud Service (IaaS)

– Compute Service: 8 and 16 OCPU Virtual Machines

CentOS, Ubuntu OS Images

RHEL via BYOI OS Image

Multipart Upload: Multipart upload enables uploading an object in parts, enhancing speed of upload and accommodating larger objects

Audit Service: This new service automatically records calls to all supported BMCS public API endpoints as log events

Search Domain DHCP

Terraform Provider: The BMCS Terraform provider is now available. Terraform is an open source infrastructure automation and management software tool

Developer Tools Enhancements: New versions of BMCS developer tools are now available, including Ruby, Python, and Java SDKs, HDFS Connector, and CLI

New Instance Shapes

– Windows BYOL: It is now possible to Bring Your Own License (BYOL) for Windows Server

– NVMe Storage: You can now use NVMe SSD disks as ephemeral data disks attached to your instances

– SSD Block Storage: These high-performance volumes can be used for persistent block storage or for bootable volumes

New Web UI: This new interface can be used to perform basic operations against Storage Cloud resources

HSM Cloud Service Integration

Oracle Exadata Cloud Machine Q&A

In Cloud, Database options, DBA, Exadata, IaaS, Oracle database, Oracle Engineered Systems on March 28, 2017 at 07:25

“The computer industry is the only industry that is more fashion-driven than women’s fashion.” Larry Ellison

Amazon have no on-premise presence, it is cloud only. Microsoft have the Azure Stack but the Azure Stack Technical Preview 3 is being made available as a Proof of Concept (POC) and must not be used as a production environment and should only be used for testing, evaluation, and demonstration.

Oracle Database Exadata Public Cloud Machine, or Exadata Cloud Machine, or ExaCM in short, is a cloud-based Oracle database subscription service available on Oracle Exadata, and deployed in the customer or partner data center behind their firewall. This allows customers to subscribe to fully functional Oracle databases on Exadata, on an OPEX driven consumption model, using agile cloud-based provisioning, while the associated Exadata infrastructure is maintained by Oracle.

This blog post contains the top 10 (5 commercial and 5 technical) facts about the Exadata Cloud Machine:

1. On-premise licenses cannot be transferred to ExaCM.

2. The minimum commitment to both the ExaCM and OCM is 4 years and the minimum configuration is Eighth Rack.

3. The subscription price for Oracle Database Exadata Cloud Machine X6 Eighth Rack is $40,000 per month (= $2,500 X 16) and that includes all DB options/features, Exadata Software and OEM DB Packs.

4. Standalone products such as Oracle Secure Backup and Oracle GoldenGate are not included in the ExaCM subscription. Only the database options (such as RAC, In-Memory, Partitioning, Active Data Guard, etc.), the database OEM packs and the Exadata storage server software are included.

5. ExaCM requires Oracle Cloud Machine to deploy Exadata Cloud Control Plane (separate subscription). OCM subscription requires similar minimum term commitment as ExaCM. If a customer already has an OCM, that can be leveraged to deploy Exadata Control Plane at no extra cost. One OCM Model 288 can manage 6 ExaCM Full Racks (i.e. 24 ExaCM Quarter Racks or 12 ExaCM Half Racks). Theoretically one OCM can support a much larger number of ExaCM full racks: about 50.

6. The 1/8th rack SKU is very similar to the on-premises 1/8th rack – i.e. minimum configuration of 16 OCPUs (cores), 240 GB RAM per database server, 144 TB raw storage (42 TB usable), 19.2TB of Flash. Compared to the Quarter Rack, it ships with less RAM, disk storage and flash. Those will be field installed if the customer chooses to go for the 1/8th to Quarter Rack upgrade. Note that this 1/8th rack enables customers to have an entry level configuration that is similar to what exists in Exadata Cloud Service.

7. Hourly Online Compute Bursting is supported with ExaCM. The commercial terms are the same as in Exadata Cloud Service – i.e. 25% premium over the Metered rate, calculated on an hourly basis. Customers can scale up or down, dynamically. Bursting does not kick in automatically based on load. Bursting of OCPUs needs to be configured by customers as needed. Once customer initiates bursting, the OCPU update is done dynamically without downtime. Customers will be billed later on the hours of bursting usage. Price: $8.401 per OCPU per hour.

8. If Cloud Control Plane is down, it doesn’t affect the availability of steady state runtime operations. However, cloud-based management (e.g. selfservice UI and REST API access) will be impacted.

9. Access: the Exadata Cloud Machine compute nodes are each configured with a Virtual Machine (VM). You have root privilege for the Exadata compute node VMs, so you can load and run additional software on the Exadata compute nodes. However, you do not have administrative access to the Exadata infrastructure components, including the physical compute node hardware, network switches, power distribution units (PDUs), integrated lights-out management (ILOM) interfaces, or the Exadata Storage Servers, which are all administered by Oracle.

10. Patching and backups: you can produce a list of available patches using the exadbcpatchmulti command as follows

# /var/opt/oracle/exapatch/exadbcpatchmulti -list_patches 
-sshkey=/home/opc/.ssh/id_rsa 
-oh=hostname1:/u01/app/oracle/product/12.2.0.1/dbhome_1

When you create a database deployment on Exadata Cloud Machine, you must choose from the following automatic backup configuration options:

– Remote Storage Only: uses remote NFS storage to store periodic full (RMAN level 0) backups and daily incremental backups, with a seven day cycle between full backups and an overall retention period of thirty days.
– None: no automatic backups are configured. Automatic backups cannot be configured later if you select the None option when you create a database deployment.

Useful links:

Oracle Exadata Cloud Machine Documentation
Oracle Database Exadata Cloud Machine Data Sheet
Features and Benefits of Oracle ExaCM
Oracle Database Exadata Cloud Machine Pricing
Creating an Exadata Cloud Machine Instance
Known Issues for Oracle Exadata Cloud Machine
Oracle SVP, Juan Loaiza, describes Oracle Database Exadata Cloud Machine

DBA Productivity and Oracle Database 12.2

In Cloud, DBA, Oracle database on February 9, 2017 at 15:15

“Technology can be our best friend, and technology can also be the biggest party pooper of our lives. It interrupts our own story, interrupts our ability to have a thought or a daydream, to imagine something wonderful, because we’re too busy bridging the walk from the cafeteria back to the office on the cell phone.” Steven Spielberg

busy

The DBA profession was recently rated as #6 among the Best Technology Jobs. Good for all of us who are in this line of business. But notice the stress level: Above Average!

DBAs are often busy people. Is that good or bad? Is “busy the new stupid”?

Automation is not a luxury for the DBAs but it is a way in which DBAs execute their job. Of course, there is one thing that cannot be automated and that is quality but the best DBAs automate almost everything.

Automating the database is a Win-Win for DBAs and DevOps. The mindset of the Enterprise DBA should be focused on harnessing the power of automation.

The following data shows what tasks are mostly and least automated:

dba_automation

Look at the last row above. I still wonder why Automatic SQL Tuning is so underestimated. It was so powerfully helping the DBA team of Nokia…

Oracle Database 12cR2 is out. And 12.2 comes with yet another new set of database automation related features:

– Oracle Data Guard now supports multiple failover targets in a fast-start failover configuration. Previous functionality allowed for only a single fast-start failover target. Multiple failover targets increase high availability by making an automatic failover more likely to occur if there is a primary outage.

– Oracle automatically synchronizes password files in Data Guard configurations: when the passwords of SYS, SYSDG, and so on, are changed, the password file at the primary database is updated and then the changes are propagated to all standby databases in the configuration.

– Online table move: nonpartitioned tables can be moved as an online operation without blocking any concurrent DML operations. A table move operation now also supports automatic index maintenance as part of the move.

– Automatic deployment of Oracle Data Guard: deployment is automatic for Oracle Data Guard physical replication between shards with Oracle Data Guard fast-start failover (automatic database failover): automatic database failover provides high availability for server, database, network, and site outages.

– Automatically set user tablespaces to read-only during upgrade: the new -T option for the parallel upgrade utility (catctl.pl) can be used to automatically set user tablespaces to read-only during an upgrade, and then back to read/write after the upgrade.

– The Oracle Trace File Analyzer (TFA) collector provides the option to automatically collect diagnostic information when TFA detects an incident.

– Oracle Data Guard support for Oracle Diagnostics Pack: this enables you to capture the performance data to the Automatic Workload Repository (AWR) for an Active Data Guard standby database and to run Automatic Database Diagnostic Monitor (ADDM) analysis on the AWR data.

– Automatic Workload Repository (AWR) support for pluggable databases: the AWR can be used in a PDB. This enables the capture and storage of performance data in the SYSAUX tablespace of the PDB.

– The new ENABLE_AUTOMATIC_MAINTENANCE_PDB initialization parameter can be used to enable or disable the running of automated maintenance tasks for all the pluggable databases (PDBs) in a multitenant container database (CDB) or for individual PDBs in a CDB.

– Automatic Data Optimization Support for In-Memory Column Store: Automatic Data Optimization (ADO) enables the automation of Information Lifecycle Management (ILM) tasks. The automated capability of ADO depends on the Heat Map feature that tracks access at the row level (aggregated to block-level statistics) and at the segment level.

– Automatic Provisioning of Kerberos Keytab for Oracle Databases: the new okcreate utility automates the registering of an Oracle database as a Kerberos service principal, creating a keytab for it, and securely copying the keytab to the database for use in Kerberos authentication.

– Role-Based Conditional Auditing: auditing for new users with the DBA role would begin automatically when they are granted the role.

– Automatic Locking of Inactive User Accounts: within a user profile, the new INACTIVE_ACCOUNT_TIME parameter controls the maximum time that an account can remain unused. The account is automatically locked if a log in does not occur in the specified number of days.

12_2_automation