Voice of the DBA

Tag: administration

  • Test Your Restores

    The ultimate testers.

    I was talking with someone the other night about their database systems and they mentioned they had implemented TDE (Transparent Data Encryption) to comply with HIPAA regulations. This person had verified that they were backing up the Database Encryption Key, which you definitely need if you want to restore a backup from a TDE encrypted database. However they weren’t sure if the certificate that protected the DEK, and the master keys on that instance were being backed up. Probably most scary to me, they hadn’t tested any restores of the database.

    Encryption is serious business, and if you are going to implement it in your databases, you had better be sure you understand how the various keys and certificates work. You better be sure you have protected your passwords, and that you can find them in the event of some issue.

    Most importantly, though, is that you need to practice recovering your database to another instance. Preferably you’d learn how to recover on an instance that hasn’t ever enabled encryption as well as one that has a different SMK or DMK.

    Practicing restores isn’t just about encryption and the potential for data loss because you don’t have a key. Practicing restores is important for all of your systems to be sure you have the skills to successfully complete a restore. It helps ensure you know where the files, tapes, disks, or any other resources are located. Most importantly it ensures that your backup process is actually running smoothly.

    Please don’t assume your backup process works. Whether you’re an accidental DBA stuck with their first SQL Server, or a ten year senior DBA that has performed hundreds of restores at previous jobs. You need to test your process and ensure that you can perform restores on the systems you are managing.

    Steve Jones

    The Voice of the DBA Podcasts

    We publish three versions of the podcast each day for you to enjoy.

  • The Default DBA

    When we setup SQL Server, there are a few choices we need to make, some of which require a few items be prepared in advance. Service accounts are one of those preparation tasks that DBAs usually perform in advance of installing an instance. There are a few things, however, that aren’t configured when SQL Server is installed, like alerts and operators. I find that often these items can be missed by inexperienced DBAs, who forget to configure standard alerts. They might also not have a mail account ready, which is required for the instance to send mail.

    Perhaps there’s a better way for SQL Server to request this information when it is installed. If one part of the setup program asked for default mail information, it would be easy to pre-configure the instance to setup alerts and enable them for communication with a DBA. Personally I’d prefer that setup get a mail profile and a default DBA group email address that receives information about the operation of the instance that the DBA should be aware of.

    There can be a hesitation from experienced DBAs that don’t want setup to install mail for them, but it would be easy to make this an optional component, with a checkbox like the one for a blank sa password. You actively choose to install the server differently if necessary, but otherwise setup would push DBAs to ensure that notifications are configured. We could even see a pre-selected list of alerts that will be enabled, with the option to disable all or just selected items.

    I don’t know that Microsoft will build these types of defaults into SQL Server, but I wish they would. We’ve done that with SQL Monitor at Red Gate, and we have seen the advantages that come when you provide good practices and knowledge to less experienced DBAs: they gain some knowledge transfer and have software running at a high level from the very beginning.

    Steve Jones

    The Voice of the DBA Podcasts

    We publish three versions of the podcast each day for you to enjoy.

  • The Last DBCC CHECKDB Date and Restores

    I ran across a question on Twitter recently where someone asked about the DBCC CHECKDB date after a restore. For those of you that don’t know this, you can run this command on your instance (in a database)

    DBCC DBINFO WITH tableresults;

    This returns a lot of information, but I’ve circled one value below:

    dbcc1

    This is the last known good DBCC date that exists for this database. It’s also the value in the error log from the execution of DBCC CHECKDB on my instance early this morning (Arrow marks the entry).

    dbcc2

    If I were to restore this database, what happens to this value? I’d expect that it would be restored to the last value that was contained in the backup file. That would make sense to me, but let’s test it.

    First I run a backup:

    BACKUP DATABASE [db1] TO  DISK = N'C:\SQLBackup\db1_20121115.bak' WITH NOFORMAT, NOINIT,  NAME = N'db1-Full Database Backup', SKIP, NOREWIND, NOUNLOAD,  STATS = 10
GO

    This should have the last DBCC date inside the file. I know run a CHECKDB and note the time in the error log.

    dbcc3

    This returns successfully (whew, didn’t want to test corruption restores). If I then run DBINFO again, I get the current date from the error log returned. Now let’s restore from my backup.

    This completes and when I run DBCC DBINFO again I find the dbccLastKnownGood date is reset back to the 12:04am value instead of the 9:11am value.

    That’s what I expect, and that was what a few other people confirmed on Twitter. It’s logical that this should be the behavior, but you never know until you’ve tested it.

  • The Cost of Storage

    SAN disks
    The cost of storage per GB drops, but the cost to companies seems to keep going up.

    One thing that we can be sure of as data professionals: we will need more storage over time. We gather more data in our organizations and more data is sent to us by partners or automated systems. At the same time we must ensure all our existing data is always available for users. I don’t know how we solve any of these problems, but I do know that any solution will require likely more storage in the future.

    As DBAs, we aren’t too often involved in the cost of storage. These days as many companies move to SAN-based storage, there are separate administrators are involved in the specification and purchasing of the actual hardware. That means that as DBAs we are removed from the impact of our requests for more storage. We don’t necessarily know what the cost to the organization is when we request another 400GB LUN. However this week I’m curious if you know the answer to this.

    What’s the cost of storage in your organization?

    I’m wondering if you know the actual cost in dollars or the impact to limited IT resources of enterprise level storage in your company. Does this impact the way you do business as a DBA? We often don’t have control over the growth of our data, but we seem to regularly be pulled into the discussions on how to best manage the amount of data and even archive data or reduce its impact on the organization’s storage.

    Let us know this week if you are impacted by the cost of storage in your organization.

    Steve Jones

    The Voice of the DBA Podcasts

    We publish three versions of the podcast each day for you to enjoy.