How user authentication works in a shared MongoDB cluster

Today, I want to clarify an important point about users authentication in a MongoDB sharded cluster because this seems to be misunderstood by some people.

In a such environment, there are two notions to understand about user authentication, especially when you are DBA:

  • cluster authentication, that is, the authentication made when connected trough a “mongos” process.
  • local authentication, that is, the authentication made when connected locally on a shard trough a “mongod” process.

Continue reading “How user authentication works in a shared MongoDB cluster”

Install EM13c agent and add targets using the Linux command line only

A few days ago, I was asked to find a way to automate the addition of newly created Linux virtual machines and databases in our EM13c console. Of course, no action should be performed via the EM web interface!

For the first part (installing the EM13c agent on newly created virtual machines), it’s quite simple, because Oracle gives you many methods (for example, “agent pull” method), in order to carry out this task. But the second part seemed more delicate according to the documentation and my constraints: everything must be done on the virtual machine. I am not allowed to connect to the EM server using SSH.

Fortunately, there is a simple solution that only requires a few extra lines in your shell script!

Continue reading “Install EM13c agent and add targets using the Linux command line only”

Elasticsearch: useful commands for DBAs

As a production DBA, you can fall into a situation where you have to act on an environment you do not know.

It can be very difficult to understand how the architecture was designed, but if you know the right commands, you can get a quick overview and all the information you need to start working.

Today, I’m going to give you these basic commands that will help you understand how an Elasticsearch installation was performed on an RHEL system.

Continue reading “Elasticsearch: useful commands for DBAs”

Retrieve the exit code of all the commands in a pipeline with “PIPESTATUS”

When developing shell scripts, it is common usage to check the exit code of a command in order to check if its execution was successful or not. It is pretty easy when you execute only one command, because the exit code can be find using the special parameter “$?”. But if you use pipelines (“command1 | command2 | …”), it is not that simple because “$?” will only give you the exit code of the last command of your pipeline.

So, today, I’m going to show you how to get the exit code for each command in a pipeline. Unfortunately, this only works in bash.

Continue reading “Retrieve the exit code of all the commands in a pipeline with “PIPESTATUS””

Recovering from loss of all control files

“Sometimes shit happens”. That’s what I thought when I realized that I had lost all the control files of one of my test databases. Of course, I had no backup because “it’s just a test database”…

Fortunately, I only lost my control files but no datafiles or redo log files. This is a very important point because if you are in this situation, all is not lost!

So today, I will show you how you can recover your Oracle database after the loss of all control files.

Continue reading “Recovering from loss of all control files”

Extract the list of all databases with server and OS version of Oracle Enterprise Manager Cloud Control repository

Today, I’m going to share with you a simple SQL query that will allow you to quickly extract useful information from your OEM repository about your databases:

  • Database name
  • Database version
  • Server name
  • Server OS
  • OS version

The query can be very useful if you need to fill in a new CMDB for example.

Continue reading “Extract the list of all databases with server and OS version of Oracle Enterprise Manager Cloud Control repository”

Get indexes usage for all collections of all databases of your MongoDB instance

One possible area of ​​improvement when you encounter bottlenecks when inserting or updating in MongoDB is to limit the number of indexes on your collections.

Even though indexes are good for queries against collections, they add an extra step when inserting or updating, which may take some time. Given this fact, it may be wise to delete unused indexes.

So, today, I’m going to give you a Mongo shell script that I use to get statistics on indexes usage for all collections on a MongoDB instance and to identify those that are not used.

Continue reading “Get indexes usage for all collections of all databases of your MongoDB instance”

MongoDB sharded cluster topology

Today, I’m going to share with you a bash script that gives you a quick overview of the topology of your MongoDB cluster.

It will give you:

  • The list and status of your mongos
  • The list and status of your config server
  • The list and status of every replica set members present in your cluster (including secondary members and arbiters), plus:
    • The date of the last role change between primary and secondary
    • The apply lag between primary and all secondary members
    • The oplog window
  • The balancer status

You can download the script on my GitHut repository “MyDBAWorld”.

Continue reading “MongoDB sharded cluster topology”

Database Link: ORA-12154: TNS:could not resolve the connect identifier specified

If you are familiar with database links on Oracle database, you may have already encountered the error “ORA-12154: TNS:could not resolve the connect identifier specified“. That is a common error which can be caused by a lot of factors. Today, I’m going to focus on the importance of environment variables to avoid this type of error message (particularly the “TNS_ADMIN” environment variable).

Continue reading “Database Link: ORA-12154: TNS:could not resolve the connect identifier specified”

Replicate a table without primary key or unique key with GoldenGate

Yesterday, I was working on a simple Golden Gate replication between two Oracle databases. Everything worked as expected, the source and target schemas were synchronized and the replication lag was about 0 seconds.

Unfortunately; after a couple of hours, I notice a 2 hours lag on the replicat

After some research on GoldenGate and with some queries on v$session and v$sql, I realize that the lag is due to a 100 million rows table on which there is no primary key or unique key defined.

Important point: I was not allowed to add constraints on the source database!

Continue reading “Replicate a table without primary key or unique key with GoldenGate”