Press "Enter" to skip to content

Category: Backups

SQL Server Backup Buffers

Published 2016-06-15 by Kevin Feasel

SQLsasquatch (whose name I now know but I like the handle so much) has a few questions about backup buffers:

I wonder:
-If ‘max server memory’ wasn’t being overridden by a more reasonable target (because max server memory is 120 GB on a 32GB VM), would the behavior still be the same before reaching target?  I bet it would be.
-Is this behavior specific to not having reached target?  Or when reaching target would backup buffers be allocated, potentially triggering a shrink of the bpool, then returned to the OS afterward requiring the bpool to grow again?
-What other allocations are returned directly to the OS rather than given back to the memory manager to add to free memory?  I bet CLR does this, too.  Really large query plans?  Nah, I bet they go back into the memory manager’s free memory.
-Does this make a big deal?  I bet it could.  Especially if a system is prone to develop persistent foreign memory among its NUMA nodes.  In most cases, it probably wouldn’t matter.

Good questions, to which I have zero answers.

Comments closed

Dueling Log Backup Jobs

Published 2016-06-14 by Kevin Feasel

Robert Davis ran into HADR_WORK_QUEUE waits recently:

Our 3rd party monitoring solution collects blocking information, but not for system threads. There was no additional information available for this blocking incident, but I could see that the system thread was a background process with the command “UNKNOWN TOKEN” and was sitting in a wait type of “HADR_WORK_QUEUE”. It was clearly the worker thread for the AG of a specific database.

A little later, we had blocking again involving that same thread, but this time, the AG worker thread was blocking the log backup thread. Seemed logical that if the worker thread could block the log backup, then the log backup could have also blocked the worker thread, but still it did not make sense to me.

This is one of those cases in which the answer makes perfect sense after the fact, but can be maddening until then.

Comments closed

Think About Recovery Mode

Published 2016-06-10 by Kevin Feasel

Kendra Little answers a user question about a co-worker who puts all prod databases into Simple recovery mode:

What if Your Coworker Regularly Goes Off the Ranch and Doesn’t Use Change Control?

Don’t cover for them.

Ask them about it first to make sure there wasn’t an emergency change request you’re unaware of for the change, but be honest about what happened when you’re asked.

In other words, treat them as an equal and a grown-up.

Sometimes in this situation, people sugar coat things or cover for the person who makes mistakes. You need to treat them as an adult though.

If you made mistakes, you would own up to what happened and work to not do it again, right? It’s just about respectfully allowing others to own their actions.

Kendra’s answer is a good mixture of technical explanation and not being socially inept.  It’s important to follow that two-pronged approach.

Comments closed

“RAID” And Backups

Published 2016-05-17 by Kevin Feasel

Kenneth Fisher explains that you can set up backup strategies similar to different RAID types:

RAID 0

Splitting the backup data between multiple files.

This is actually a fairly common way to speed up a large backup. IO is one of the slowest part of the whole process so by splitting the backup into multiple files we can reduce our backup time by quite a bit. Having multiple files will also increase your restore time but you do have to make sure that all of the files are available. If you lose one file then the whole backup is useless.

This is done by listing multiple locations for the backup.

This is an interesting use of RAID as metaphor.

Comments closed

Restoring Azure SQL Databases With Powershell

Published 2016-05-12 by Kevin Feasel

Mike Fal shows us how to restore an Azure SQL Database database using Powershell:

The most fundamental form of disaster recovery is database backups and restores. Typically setting up backups is a lot of work. DBAs need to make sure there’s enough storage available for backups, create schedules that accommodate business operations and support RTOs and RPOs, and implement jobs that execute backups according to those schedules. On top of that, there is all the work that has to be done when backups fail and making sure disk capacity is always large enough. There is a huge investment that must be made, but it is a necessary one, as losing a database can spell death for a company.

This is one of the HUGE strengths of Azure SQL Database. Since it a service offering, Microsoft has already built out the backup infrastructure for you. All that stuff we talked about in the previous paragraph? If you use Azure SQL Database, you do not have to do any of it. At all.

What DBAs still need to manage is being able to restore databases if something happens. This is where Powershell comes into play. While we can definitely perform these actions using the portal, it involves a lot of clicking and navigation. I would much rather run a single command to run my restore.

The Powershell cmdlets are easy to use, so spin up an instance and give it a try.

Comments closed

Restoration Options

Published 2016-04-21 by Kevin Feasel

Richie Lee covers the WITH RECOVERY, WITH NORECOVERY, and WITH STANDBY database restoration options:

Similar to NORECOVERY except that the database will accept read only connections. To do this any uncommitted transactions in the backup will be rolled back and stored in a transaction undo file (tuf.) Whilst users are running queries against the database no further restores can continue until all queries are complete (though this is not the case with log shipping.) When the next restore occurs, those uncommitted transactions in the tuf file will be rolled forward and the next log is restored.

Diving into STANDBY mode was quite helpful.  I’ve never needed to restore a database into standby mode, but could see it being useful for bringing back deleted records.

Comments closed

SAN Snapshots Aren’t Backups

Published 2016-04-01 by Kevin Feasel

Denny Cherry explains why SAN snapshots aren’t a good backup policy:

SAN snapshots, and I don’t care who your vendor is, by definition depend on the production LUN. We’ll that’s the production data.

That’s it. That’s all I’ve got. If that production LUN fails for some reason, or becomes corrupt (which sort of happens a lot) then the snapshot is also corrupt. And if the snapshot is corrupt, then your backup is corrupt. Then it’s game over.

SAN snapshots are a good part of an infrastructure-side recovery plan, but databases have their own recovery time and recovery point objectives.  Conflating these two can lead to sadness.

Comments closed

Availability Group Backup Failures

Published 2016-03-31 by Kevin Feasel

James Anderson had a recent experience in which a database in an Availability Group failed to back up properly:

Last week, I received an alert that the percentage of transaction log in use on one of our production databases was increasing more than it should have been. I’ll refer to this database as DB1 from here on in. It had reached 18%, which is above the normal 5-10% we like to see it at. DB1 is in an Always On Availability Group so if the log usage jumps we jump.

The first thing I checked was that the log backups were running. We use Ola Hallengren’s maintenance solution to manage our backups. The backup jobs were running without error.

The next thing to check was the log_reuse_wait_desc column in sys.databases.

The highlighted row is the row for DB1. A log_reuse_wait of 2 means there were no changes to backup since the last log backup. If that were the case then why was my transaction log slowly filling like an unattended bath?

Read on to learn more.  James also has a link to a (closed/won’t-fix) Connect item.

Comments closed

Restoration Failures

Published 2016-03-17 by Kevin Feasel

Tibor Karaszi shows us cases in which Management Studio can generate an invalid database restoration sequence:

Above, the GUI incorrectly base the restore on a copy only backup. After using the timeline dialog to point to an earlier point in time, you can see that the GUI now has changed so it bases the restore on this potentially non-existing copy only backup. Not a nice situation to be in if the person doing the restore hasn’t practiced using the T-SQL RESTORE commands.

It’s important to be able to write the relevant T-SQL queries to restore your database, just in case you run into one of these issues.

Comments closed

Restoring A Striped Backup

Published 2016-03-11 by Kevin Feasel

Steve Jones shows us how to restore a backup striped across multiple files:

Each of these is part of a striped backup, a piece of a backup file. To restore the backup, I need all the files to reassemble the backup. This is fairly simple, but you should be aware of how this works and how to perform a restore.

In my case, you can see I have 7 files for each database. They are the same name with an _0x at the end, with x being the number of the file.

Take Steve’s advice and script out the restore process.  That way you know how to do it next time, can start building automation scripts, and can make your life easier.

Comments closed