Robin Day's questions

My AWS account runs approximately 50 instances.

They all run Windows, but they vary between being just Windows, or running SQL Server STD or SQL Server Web.

I'd like to find a way of exporting all my instances along with their Instance Type (r3.xlarge, etc) but also the SQL Server Licence they are running.

This is to enable me to pull this data into various reporting tools so I can compare the running cost per hour of each server against clients, etc.

At the moment, I can pull and automate the Instance Types, but I have to manually enter the SQL Server licencing in order to get the cost per hour.

We have a web application running on an EC2 instance that was built using the SQL Server 2008 R2 Standard AMI.

We have since noticed that the cost of Standard is substantially higher than the cost for Web.

I cannot find any reason to indicate why it would not be acceptable to run under the Web licence of SQL Server.

Is there any way to modify / update the licencing on the stance to go to Web from Std?

I am hoping there is a simple setting / key thing that can be done, rather than having to create a new instance with the Web licence and re-installing.

I have SQL Server 2008 standard edition running on Windows Server 2003.

I need to encrypt the databases.

One obvious option is to upgrade to enterprise edition and use TDE.

However, this might not be an option due to cost.

Can anyone provide any information about if it is possible to use Windows Encrypting File System to encrypt the SQL Server Data directories?

Would this work? Would this cause any major performance hit?

Any other alternatives would also be greatly appreciated.

I have a production database with Full recovery mode where the MDF and LDF files are approx 80GB each.

We have a test server that we restore this database to on a regular basis for dev / test purposes, etc.

At the moment the restore has happened, we set the recovery model to Simple and shrink the log file as the server is low on disk space.

Is there a way of restoring the database straight to Simple mode so that the large LDF file is never created?

I am running some test scenarios with setting up a server farm using Windows 2008R2, IIS7.5, WebFarm Framework and Application Request Routing.

I have three servers:

WebFarm1    Controller    192.168.0.231
WebFarm2    Primary       192.168.0.232
WebFarm3    Secondary     192.168.0.233

Using just the default site on all three servers I can successfully browser to http://192.168.0.231 and it will load balance between WebFarm2 and WebFarm3.

I would like to setup the WebFarm so that it can replace an existing server that runs multiple sites with different bindings. For example:

TestServer1    TestSite1    192.168.0.241    D:\Sites\TestSite1\
TestServer1    TestSite2    192.168.0.242    D:\Sites\TestSite2\
TestServer1    TestSite3    192.168.0.243    D:\Sites\TestSite3\

How would I configure the WebFarm to do this?

• Which Server do I add the three IP addresses to? I would guess at WebFarm1
• I can create the three sites on WebFarm2 but how can I set up the bindings if it doesn't have the IP addresses?
• Do I need to create the three sites on WebFarm1 as well?

Can anyone help or point me somewhere with any articles?

We had a virtual machine running on a VMWare Server host on Windows Server 2003.

The machine is set up with non persistent disks.

We had a windows task schedule set up that ran a batch file to reset the machine each week so that it returned to it's original state.

The batch file that we had running was:

"C:\Program Files\VMware\VMware Server\vmware-cmd" "C:\Virtual Machines\VirtualMachineName\VirtualMachineName.vmx" stop hard
"C:\Program Files\VMware\VMware Server\vmware-cmd" "C:\Virtual Machines\VirtualMachineName\VirtualMachineName1.vmx" start

We have since migrated this machine to the free version of ESXi 4.1.

Can anyone let me know if and how it is possible to schedule such a restart?

I've installed Ola Hallengrens Maintenance scripts and it has created the jobs for me.

DatabaseBackup - SYSTEM_DATABASES - FULL
DatabaseBackup - USER_DATABASES - FULL
DatabaseBackup - USER_DATABASES - DIFF
DatabaseBackup - USER_DATABASES - LOG
DatabaseIntegrityCheck - SYSTEM_DATABASES
DatabaseIntegrityCheck - USER_DATABASES
IndexOptimize - USER_DATABASES

I plan on following his guidelines to the question on his FAQ How should I schedule the jobs?

This depends on your maintenance window, the size of the databases, the maximum data loss and many other things. Here are some guidelines that you can start with, but you will need to adjust it to your environment.

User databases: Full backup one day a week. Differential backup all other days of the week. Transaction log backup every hour. Integrity check one day a week. Index optimization one day a week.

System databases: Full backup every day. Integrity check one day a week.

Integrity check after index optimization. This is because index rebuilds sometimes can fix database corruption. Full backup after index optimization. Then the following differential backups will be small. Full backup after the integrity check. Then you know that the integrity of the backup is ok. This means first index optimization, then integrity check and finally full backup.

My question is still, How should I schedule the jobs?.

In particular:

If I perform a full/diff backup at midnight each day, should I run a transaction log backup at midnight as well? Or, should I make the midnight job run a transaction log back and THEN do a full/diff backup? Or, do I just not perform a transaction log backup at midnight?

How should I setup a job to perform the index optimization, then itegrity check and then the full backup? I don't want the diff backups and transaction log backups to be huge after an index rebuild unless absolutely necessary.

Any advice as to how others have set this up would be great.

I need to send a client an SSL Certificate Request (from IIS 6.0)

They have requested that it ONLY be sent encrypted using an app like Winzip.

My question is. Is this actually an issue?

My very basic understanding of certificates is that the request does not contain any information that would help a hacker in obtaining the final certificate.

Am I completely wrong?

I have to add a new mime type to an IIS 6.0 web server.

The client would like this done ASAP as it currently isn't serving .flv files.

However, the server is used for other clients and I would like to avoid doing an IIS Reset during working hours.

Does anyone know a way of getting IIS to read the new mime types without a reset?

I'm looking at virtualising some of our infrastructure in order to allow for more resiliance and future expandability.

We have successfully virtualised on single servers with Direct Attached Storage and are now looking for a more future proof solution using a high powered host (or two) and a SAN (or two).

I'm thinking that the host machine will probably be an HP ProLiant DL360 G7 (all of our exisiting infrastructure is HP).

Unfortunately, I am new to the world of SANs. From what I can see, the Buffalo Terrastation III is all I would need in order to setup an iSCSI SAN for VMWare to use. However, I'm a little reticent to go that way as it's a bit too "entry level" for my liking. In particular I would be very keen for more redundancy, power, networking, etc. I'm also very aware that you "get what you pay for".

Therefore, can anyone reccommend equivalents from the big boys? HP/IBM? I have searched high and low on the HP site and seen many options but am struggling to work out if it is all the hardware I will need. Some options appear to need separate controllers from disk enclosures, etc.

We run a number of web apps that store a lot of local data in small xml files. One part of our backup / recovery strategy is to produce a local mirror of the file system via a VPN to the hosting centre.

The VPN connection is only via a 12Mbps ADSL and whilst there are a lot of files and directories, the actual number of files that changes is quite small.

Although the bandwidth is probably an issue, I'm seeing results such as the output below. The robocopy /MIR took 5 hours to run yet only 30 mins to actually perform the copy.

Does anyone have any suggestions as to ways to improve this. The 5 hours is now bordering on too slow and if we can't find a way to speed this up then we're going to have to come up with a completely different solution.

            Total    Copied   Skipped  Mismatch    FAILED    Extras
 Dirs :     17625      6618     11007         0         0         0
Files :   1112430      1223   1111207         0         0         0
Bytes :  57.451 g  192.25 m  57.263 g         0         0         0
Times :   5:01:23   0:35:55                       0:00:00   4:25:27

Speed :               93509 Bytes/sec.
Speed :               5.350 MegaBytes/min.

Ended : Fri Apr 16 05:54:23 2010

We are a small business that hosts our core product on a number of HP servers.

Our core production setup is

• 1x HP DL380, high powered for a SQL Server Database
• 1x HP DL360, mid powered for our core application server
• 6x HP DL320, low powered for our front ends

We run our training / testing / support systems on a similar setup, the servers are just older and less powerful. Unfortunately this is now causing us issues as the system has grown beyond the capabilities of these older servers.

Upgrading these servers would be expensive and we believe that virtualisation is probably the way to go for the future. Locally we run a number of test / dev environments on ESXi using Direct Storoage on a couple of high powered DL360's and these are performing fairly well.

We're thinking that instead of replacing all of our test servers that we can implement an iSCSI SAN and one or two high powered hosts. Hopefully looking that when it comes to replace our live servers as well that we can just expand the virual environment to cope.

So my question is... Can anyone offer any advice on some suitable options? We have generally always been extremely happy with HP servers, all of our kit is currently HP, therefore our preference would be to stick with HP, however, I'm always happy to hear about other options.

I'm hoping that initially a budget of around 15-25k (GBP) would be suitable, this could potentially be increased if I had confidence that the system would pave the way for a cost effective upgrade of our live systems in the future as well.

I am new to SAN's and my only real experience is playing with OpenFiler on some old desktops. I think iSCSI should be suitable, but I've not done any research into how SQL server may perform. I've had a browser through HP's sites and see plenty of information about EVA, MSA, LeftHand, etc. However, from looking at all that, I don't see which options would be best and more importantly I don't know exactly what I would need to buy.

Any help, links, opinions would be much appreciated.

Thanks

Possible Duplicate:
How to sleep in a batch file?

I have a batch file that runs as a secheduled task in windows server 2003.

This batch file purely calls around 20 other batchfiles one after the other.

When this runs it causes quite a spike on the server and a reduction in performance.

I'd like the ability to add a WAIT or SLEEP type command between each of my batch file calls in order to spread the load out a bit.

Can anyone shed any light on the simplest way of doing this?

I have a Virtual Machine running Windows 2003 Server which itself is running on VMWare Server 1.0.4.

The Virtual Machine has Non-Persistent Disks and periodically I manually shut down the box and turn it back on in order to put it back to it's original state.

Can anyone tell me how I can schedule this to happen on a weekly basis?

On of our HP Proliant DL320 G4's is reporting a failure. It's running Windows 2003 and configured using the built in software SATA Raid.

The array manager shows only one drive attached labelled "Port 1" and so it's in a degraded state.

I have the replacement drive sitting on my desk and I'm all ready to go to the hosting centre to swap it out... My question is... how do I know which drive to pull?

The on board software raid on the HP's does NOT show any lights on the disks and therefore it is not possible to see which one has failed, which one is working or the ability to blink a specific drive.

Is there a simple rule that Port 0 is on the left and Port 1 is on the right?

If you right click the network icon in the system tray in Windows Server 2003, right next to "Status" is "Disable".

If you "misclick" this, the result can be catastrophic. Is there a way to disable it, or remove it from the list to avoid "accidently" clicking it?