Condusiv Technologies Blog

Condusiv Technologies Blog

Blogging @Condusiv

The Condusiv blog shares insight into the issues surrounding system and application performance—and how I/O optimization software is breaking new ground in solving those issues.

Best Practices for CSV defrag in Hyper-V (Windows Server 2008R2)

by Michael 28. March 2011 04:33

One of the most significant features in Windows 2008R2 (for Hyper-V) is Cluster Shared Volumes (CSV) for virtual disks (vhd). This allows NTFS to behave similar to a clustered file system, addressing many limitations found in Hyper-V storage with the original release (Windows 2008).  

There are three online modes/states for CSV:
  • Direct Access: In this state, the CSV is available to all nodes in the cluster (i.e. all your VMs) for direct high performance storage access. This is the state you want in production.  
  • Redirected Access: In this state, the CSV is still available to all nodes in the cluster, but all I/O is redirected through a single "coordinator" node. Redirected access is used in planned situations where you need to perform certain disk actions that can't have multiple nodes accessing and locking files concurrently, such as a VSS backup or defrag. Channeling all I/O through a coordinator slows I/O and is more likely to cause bottlenecks for production demands.
  • Maintenance mode: enabling this mode is a safe means to get to a state where processes that require exclusive access to a volume can be used, such as a maintenance routine like chkdsk.

Best Practice: 

  • On the Hyper-V system volume,  pass-through volumes and any other non-CSV volumes, leave Automatic Defragmentation on at all times.
  • Given the performance benefits of Direct Access for cluster shared volumes, leave IntelliWrite on and run an occasional scheduled defrag. This is because of the requirement to use the coordinator node and place the volume into a Redirect Access state. Automatically changing from direct to redirect and back is all part of the file system control (kernel code we co-wrote with MS in the mid 90’s – as a Windows source code licensee), and the mechanism all defragmenters use today - you do not need to do anything special.
  • Correction (June 30, 2011): In the process of testing for the V-locity 3.0 release, we discovered that defagmentation does NOT cause a state change to Redirected Access. This is true for any defragmenter. So, defragment CSVs as you would any other volume. [Apologies on making this statement without validation - we should know better :-)] 

Diskeeper and V-locity are fully compatible with CSVs as confirmed by Windows IT Pro here. The file system control built into Windows is used to defrag, but not used for prevention in the design of IntelliWrite, which is a CSV-compatible file system filter driver (it's very important for drivers to be CSV-compatible) residing at a low altitude, expect on XP (where its altitude is much higher). You can view all file system minifilters and their allocated altitudes here.

IntelliWrite is “DKRtWrt” (its code names in development stages was WriteRight and then later RightWrite -hence "RtWrt"). To see or load/unload filter drivers, use the Filter Manager Control Program (fltmc):

Tags: , , , ,

Defrag | Hyper-V | IntelliWrite | V-Locity

Faster Backups/Archiving/Dedupe/DR success with Diskeeper and V-locity

by Colleen Toumayan 27. January 2011 03:29

"Spokane Regional Health District uses CommVault Simpana backup/archiving/disaster recovery software installed on a dedicated server with 37TB of SAS attached storage.

                                                                                       

We perform daily full and incremental backups of all our servers. The data backup is disk-to-disk-to-tape and is deduplicated as it is saved on the SAS storage. The deduplication process can create a very large number of file fragments, sometimes over 1,540,000 fragments on a 2TB disk array. With Diskeeper EnterpriseServer automatic defrag running the response time of the arrays is approaching 0.02 second delay due to fragmentation. This has reduced our backup time by approximately 25 percent for any D2D2T job. 

SRHD also uses Microsoft Hyper-V and currently has 31 virtualized servers running on an Intel Modular Server. There are 72TB of storage available to the Modular Server via SAS connections featuring dual path IO. All of the data on the SAS arrays is maintained in RAID 60 logical disk drives. Since setting up V-locity, which has built-in support for VHD (virtual hard disks), with automatic defragmentation, our VHDs very seldom show any fragmentation. 

                                                         

The solutions also have the intelligence to monitor disk IO and the defragmentation will pause to prevent IO latency affecting performance. They are set and forget applications which perform a very well without impact on our server response times."

-Larry Smith, Spokane Regional Health District

Tags:

Defrag | Diskeeper | SAN | V-Locity

Defragmenting IT Healthcare

by Michael 20. December 2010 05:18

Joe Marion is founder and Principal of Healthcare Integration Strategies, specializing in the integration of imaging technologies with the overall healthcare IT landscape. His blog (at Healthcare Informatics) covers challenges and opportunities specifically relevant to optimizing Healthcare IT initiatives.

Medical images are a significant percentage of the the world's storage requirements, and have been predicted to encompass an even greater percentage of future storage demand. In Joe's recent blog post he posed the question "Is Defragmentation a Boon to Healthcare IT Performance?"

In his post he includes personal observations and insight into performance implications fragmentation can incur on IT as healthcare departments themselves consolidate and standardize application use:

"With departmental solutions, there very likely was less emphasis on system tools such as defragmentation applications.  Now that PACS technology is becoming more intertwined with the rest of IT, there should be greater emphasis on inclusion of these tools.  In addition, server virtualization can mean that previously independent applications are now part of a virtual server farm."

He also makes the astute observation that centralizing computing and storage magnifies bottlenecks, making a solution such as defragmentation increasingly more vital:

"The addition of disk-intensive applications such as speech recognition and imaging could potentially impact the overall performance of these applications.  As data storage requirements within healthcare grow, the problem will potentially get worse.  Think of the consequence of managing multiple 3000-slice CT studies and performing multiple 3D analyses.  As more advanced visualization applications go the client-server route, the performance of a central server doing the 3D processing could be significantly impacted."

You can read Joe's blog here.

  

Tags: , , ,

Defrag | Diskeeper | IntelliWrite | V-Locity

Storage VMotion and GOS fragmentation

by Michael 3. December 2010 06:57

I had a test run here internally in order to make a point about what does, or more specifically "does not", happen when you VMotion/SVMotion a Windows Guest OS (GOS). We wanted to demonstrate that, while VMware is copying the VM to another host/storage, it does nothing about the internal fragmentation of files in Windows.

We felt this was a valuable demonstration as one of the old (1980s) ways to "fix" fragmentation was to copy off the files/backup, reformat the volume, and then copy back/restore. This offered a degree of success, but required taking the data offline in order to get rid of most of the fragmentation. On a side note, backing up/copying fragmented files takes a lot longer than it would on contiguous and ordered files.

Anyway, S/VMotion is such a cool feature because it works on live VMs. So, if the VMDK movement somehow did align/reorder files in Windows, it could be a great solution to Windows file system fragmentation! So here's how we tested...

1. Setup 2 ESX 4.1 Servers with iSCSI storage and vCenter with SVMotion capability.

2. Create a VM with Windows 7 in one of the ESX Server storage (Ex: Storage1) and a 20 GB Thin virtual disk.

3. Using an internal tool, create moderate fragmentation on the virtual disk (80k fragments, average fragments per file around 3.0, around 50% free space).

4. Install V-locity with all features (e.g. defrag, IntelliWrite, etc...) disabled. This is just so we can run a fragmentation analysis and save the reports.

5. Save the "Before SVMotion" analysis report, and then stop V-locity Windows Service (to make sure it is entirely inactive).

6. Using SVMotion move the live VM to the other ESX Server storage (Ex: Storage2).

7. Once the move is completed, restart the V-locity Windows Service and perform a post "After SVMotion" analysis.

8. Save this job report.

We saw what we expected, given VMotion leverages Changed Block Tracking (CBT) technology and is block, not file based. I attached the report, so you can see the side-by-side analysis data, files in Windows are not defragmented in an SVMotion. Now, that's not to say possible fragmentation of the VMDK files themsleves (on VMFS datastores) was not affected, but that's a topic for another post. 

V-locity is VMware Ready

by Michael 12. November 2010 06:16

As a long term VMware Premier Technology Alliance Partner, we're excited to now announce that V-locity has, after passing stringent testing, been certified by VMware as VMware Ready. V-locity is the first and only certified solution for virtual disk I/O optimization. 

You can see the product listed at VMware's site here.

  

Tags: , , ,

V-Locity | VMware

Month List

Calendar

<<  November 2017  >>
MoTuWeThFrSaSu
303112345
6789101112
13141516171819
20212223242526
27282930123
45678910

View posts in large calendar