Outlook 2007 Mutliple Accounts

We have someone here set up with multiple accounts in Outlook 2007. We don't have an exchange server. When this person sends a meeting request from the non default account, it says it's from [non default email] on behalf of [default email].

My temporary (maybe permanent) solution is to have him switch his default account whenever he needs to send out a meeting request.

Is there anyway to stop this?

From serverfault $pageUsers[$entry.posts[0].author].name

• The way we have solved this is to create another E-mail account under Account Settings. When you do that, a drop down will appear under the Send button when creating a new email. It will say Account. Change the checkmark to the appropriate account.

  From Joseph

Cleaning Thunderbird Inbox of old viruses

I am running BitDefender and it seems to be picking up stuff that Norton never mentioned. The problem is that BD reports a number of old viruses in my inbox that it says it can't disinfect because they've been archived.

I assume these are all messages that have been deleted (some go back to 2006). So I thought compressing my inbox would get rid of them. But so far BD is still reporting them.

To make things even more confusing, Thunderbird seems to be using both my Application Data/Thunderbird and Application Data/Mozilla (the directory I used before, and which I told Thunderbird to keep using). Both of which were updated today!

Since there are a few hundred of these messages, I would really like to clean things up.

Oh, and my two inboxes are too big to edit! Ideas, tools, etc., would be appreciated, as I'm not sure which questions to ask.

From serverfault Paul Morrison

• Classic case for the Tiresome Manual Binary Chop pattern. Remember to back up first. Complexity is something like O(m log n), with m the number of viruses and n the number of messages. You can probably speed things up a bit by dealing only with messages that have attachments.

  Avery Payne : You mean this? http://en.wikipedia.org/wiki/Binary_search

  Paul Morrison : What is the relevance of this answer? What tool can I use to clean up the Inbox(es)? Sorry to be dumb!

  Avery Payne : He's saying "make a backup, sort your stuff by hand for attachments to create a primitive filter, then go over only the stuff you've 'filtered' to cut down on the time involved", if I'm understanding this correctly.
  From $pageUsers[$post.author].name

• When you say "compressing your inbox", do you mean "compacting your folders"? You should definitely compact your Inbox if you haven't done so already. You can even set Thunderbird to do so automatically Every Once In A While (Preferences > Advanced > Network & Disk Space > "Compact folders when …" — why this isn't already configured this way, I'm not sure). This might not solve your problem, but it's hopefully a starting point.

  Paul Morrison : Thanks Paul - it was switched off for some reason. Any idea what the previous answer means?!

  Paul Fisher : Sorry for not getting back sooner. Thunderbird doesn't compact folders at all by default, and I'm not sure why.
  From Paul Fisher

Subversion Client for Windows Vista

Can someone recommend me, free if possible, subversion client for Vista?

From serverfault Sergio del Amo

• tortoiseSVN is very good.

  GreenKiwi : If you find you are having performance problems, you can disable caching in TortoiseSVN and that can improve machine performance.

  squillman : Aye, is the way to go! +1

  Brian Knoblauch : Only downside is that they release updates frequently, and they require machine reboots (it's exceptionally rare for me to reboot my Vista dev box, other than windows updates, tortoise updates, and the occasional BSoD).

  GreenKiwi : Actually, I find that the reboots aren't really needed. Particularly on the updates.

  Ryan Bolger : Just because there are frequent updates doesn't mean you have to apply them. If there's nothing changed that you need, just skip it. I much prefer frequent updates to glacial updates though. Active projects are good.
  From Garry Harthill

• SubClipse is great if you are using Eclipse.

  CoverosGene : update site: http://subclipse.tigris.org/update_1.6.x Subversive (http://www.eclipse.org/subversive/) is another Eclipse option.

  Jon St. John : Thanks - guess since I'm new I can't post links yet.
  From Jon St. John

• There is always the trusty command line version of svn. I use this in addition to GUI versions, I have run into too many instances where the GUI version has a problem and I have to resort to the command line to fix it.

  I have also used SmartSVN and been very happy with it. It is more of a full fledged client, rather than an extension to the explorer. This has some benefits, in that it will monitor and show you all your current changes and those pending changes.

  From GreenKiwi

• I really like VisualSVN if you're using Visual Studio. It's free for a 30 day trial, then $49. Well worth the 50 bucks, in my opinion.

  If you have to have free, and using Visual Studio, try AnkhSVN. I've used it before, and it works fine.

  Redbeard 0x0A : +1 for the VisualSVN client for Visual Studio. Excellent product, even their *FREE* server product is awesome.

  Brian Knoblauch : I've heard good things about AnkhSVN, but I couldn't get it working on my system.

  Bert Huijben : @Brian: At the AnkhSVN support mailing list we don't have any open installation Issues. Can you tell us about your problem? (Not much we can do if we don't hear of our (possible) users :))
  From Nathan DeWitt

• Subcommander has been very helpful for me; it too is a full-on client as opposed to an Explorer extension. The GUI is a little non-inutitive at first, but it works great otherwise.

  From yalestar

• Probably not the answer you look for, but the Subversion does include a client when you install it: the svn command. It is an excellent client when you get used to it. Combined with a merge tool for handling conflicts you have all the power from your keyboard.

  From tronda

• Tortoise is great. Having both is not a problem, and many environments, such as Netbeans, will take full advantage of the command line interface if you have it.

  From Philip

• I use SlikSVN on Windows 7 (So I imagine it'd work fine on Vista). It's a command line client, and does exactly the same as the unix SVN. I also use TortoiseSVN for gui goodness - there's no reason you can't use both.

  From Ian Rolfe

What steps do I need to perform to host a ASP.NET web service publicly?

I am really a programmer and new to web development field. I am tasked with creating a web service which must be publicly hosted (on internet) and which one of your application will use when needed.

I have developed the web service (using ASP.NET/WCF) and have tested it by hosting it on one of the server in INTRANET. The service uses a database which MUST be installed in organisation intranet only (some sensitive data). I am now confused as how to proceed to host it on internet. Please can you guide about the steps I need to perform?

From serverfault Hemant

• If it's been tested and you're happy with it working then you need to work with your security folk to have connections to it allowed through the firewall on HTTP / HTTPS.

  If you're doubling as the security folk, then what I would recommend is to put it in a DMZ off of your firewall, keeping the database server inside the firewall. You would then need to create an incoming path for the database connections from your DMZ web server (that hosts the web service) through the internal interface of the firewall to your db server.

  If you post more details regarding the technologies you're using (MySQL / MS SQL, are you the security person, etc) then more details can be provided on how to implement this.

  Hemant : I am using SQL Server 2008 Express edition as database server. I am not a security person or network administrator so honestly, I didnt understand the term DMZ. Please can you elaborate a bit?

  Hemant : Oh! and thanks for the earlier answer :)

  squillman : Ok, no problem. Can you please edit your question to include a little more information regarding your network setup? What are you using for a firewall? Will you be the one making the configuration change to allow access to the web service?
  From squillman

• If you have successfully tested it on the intranet of your organisation then to expose it to the internet is relatively straight forward, you just need host the application on a publicy accessible server.

  It is common to have publicly accessible site use a database back end which is hosted securely inside hte organisations firewall and have only the application access this database through a secure channel, therefore exposing as little as possible.

  So basically to expose an application on the internet, all you need to is deploy it on an Internet accessible server, the same way as you hav done on the internet. Obviously as this is now available to all you need to make sure your site is secure and not going to expose the database to attack.

  From Sam Cogan

Completely remove .Net 3.0 install

Possible Duplicate:
What do I do if .NET Framework is not installing?

Trying to install .Net framekwork 3.5 but it complians that 3.0 is not present.

However, when trying to install 3.0: the installer complains that 3.0 is in fact installed.

• Add/Remove programs doesn't list 3.0 (When I tried to uninstall it, I got a message saying the install program coulnd't be found and did I want to remove it from Add/Remove programs, I said yes).
• I'm sure this is registry problem but I can't find any associated keys.

Any ideas?

Thanks

EDIT: Solved I ran the ".NET cleanup tool" found from this question: What do I do if .NET Framework is not installing?

So, since this question is essentially a duplicate, shoudl I delete it or wait for an admin to close it?

From serverfault cbrulak

• Are you running the upgrade for the .NET Framework?

  EDIT
  Try running the .NET cleanup tool.

  RSolberg : This is the answer from your post. No reason to delete the question as its always good to have a different question that leads you to the answer.

  Kara Marfia : Agreed. Tho it'll probably save someone some time if you notch an answer.
  From RSolberg

• As the 3.5 SP1 installer contains 2.0 and 3.0 that shouldn't be needed anyway... have you tried the full offline setup instead of the bootstrapper (which only downloads what it thinks you need)?

  cbrulak : yes I have, but no luck
  From Oskar Duveborn

SQL Server 2005: Copy SQL Server Agent Job?

I have a SQL Server Agent job comprised of 10 steps and each step has many parameters defined. Now I need another job almost exactly the same, but with a few differences. Since "copy / paste" isn't an option for jobs, I'm looking for suggestions on what would be the best way to create a new job based on the first.

Edit: I just noticed that I can right-click and "Script Job As". I suppose one option would be to do just that and edit the script for my changes. I'm still curious as to if there are any other (better?) options as well.

From serverfault Chris_K

• I don't think there is a better option than scripting your job. You can modify the script as required and reapply it to any number of SQL instances (always backup your system databases before apply scripts).

  Chris_K : This seems to be the best way to go when copying jobs in the same instance.
  From SuperCoolMoss

• You may find the SQL Server Integration Services Task, Copy Job to be of use.

  http://msdn.microsoft.com/en-us/library/ms137568(SQL.90).aspx

  Chris_K : I had looked at that, but it doesn't seem to help me copy a task on the same instance. I guess I could copy from instance a to instance b, rename and modify and then copy back but with my ADD issues I'm sure to lose track of what I'm doing halfway through! :-)
  From John Sansom

Recover Outlook PST files

I have a 6 GB Outlook file that won't load in Outlook any more. I am not sure what recovery options I have available. I downloaded one recovery tool demo which shows some but not all mail messages.

Are there any tools that have worked for you? Can I import the PST files directly into another software?

EDIT: I have tried scanpst quite a few times but it fails to recover the file.

From serverfault aleemb

• See the official Microsoft page for some more info on this and a possible solution using Outlook 2007:

  http://support.microsoft.com/kb/932086

  There's also a bunch of free/commercial tools for splitting PST files, e.g:

  http://www.splitpstfile.com/

  aleemb : The PST file splitter should come in very handy! At least now I can divide and conquer.
  From Jon

• Have you tried SCANPST.exe - its part of the office install and is used to recover corrupt PST's Sometimes known as the inbox repair tool The common install path is

  C:\Program Files\Microsoft Office\Office12

  From UndertheFold

• Always try ScanPst.exe first, but it doesn't always mange to recover the PST.

  At this point you are left with a vast array of commercial options starting at around $70, I've had success with recoverytoolbox.com

  But your experience will obviously vary. The recovery process will take a LONG time (5-10 hours) due to the size of your PST file - in the future you may consider archiving it to make maintenance on the file quicker!

  PS Since 2003 the maximum size for a .pst is 20GB

  From Jon Rhoades

• Here you must use PST File Splitter Software to first split that 6 GB file into small files & then start the recovery process using some efficient PST Repair Software. You can opt for following softwares:-

  1) http://www.split-pst-file.com/

  From $pageUsers[$post.author].name

How can I configure an USB Stick under Vista so that I can remove it while in-use without data loss?

Hi,

I've got an USB-stick with my keys and passwords on it and I'd like to remove it from the PC when I leave my desk, but without having to check the filesystem on it every time as it happens if I do it now.

Is there any possibility to improve the situation under Vista?

From serverfault David Schmitt

• Personally I have never had a problem with data loss in XP or Vista when simply pulling a USB stick out. The key factor is to make sure data is not being written to the disk when removing it.

  I do reformat my USB devices to use NTFS tho, (which they typically are formated as FAT32 from the factory).

  Alternately you will see an icon near your PC clock to safely remove a device from the system. Simply double click and select USB Mass Storage Devices and select the device by drive letter/model. Then pull the device.

  (Edit to add): Optionally, make sure you have the device set to "Optimize for Quick Removal" in the device properties, under the device manager. This turns off caching of files (which is usually the default setting for removable devices).

  Mark Brackett : +1 to Optimize for Quick Removal. Otherwise, writes are cached *and you will be burned* if you don't religiously use Safe Removal.

  Jack B Nimble : I have had jump drives become completely corrupted by not using safe removal, but it is rare.

  : Not saying it is impossible to corrupt a USB drive by just pulling it...but with the various and numerous drives I have used, I usually wash them before getting a corruption (well, I've washed 4 so far (only one died...probably due to drying it too :( ), but have never had a unusable disk from a pull. YMMV.

  David Schmitt : Thanks, the "Optimize for Quick Removal" sounds like what I was searching for. The strange thing is that the option is disabled on my thumb drive, but there is an extra checkbox to disable write caching. I'll try that one.
  From

• Does your USB drive have a read-only switch? Make it read-only, if you don't try to write anything, then you should be able to pull it out safely without any loss. If you don't have a switch, check the drive manufacturer's site, some have a software tool that allows you to mark the drive as read-only.

  From Zoredache

Is there any way to find out how many hits you receive from bots, on an IIS6 site?

Hi folks,

Is there any way to figure out how many bots (even if it's just one type of bot) that hits our site, on a daily, weekly and monthly average?

we have the standard IIS6 web site logs in lots of files.

Any tricks or recommendations?

Update

Program needs to be .NET based. No php or extra ugly packages, please. Free would be a bonus.

From serverfault Pure.Krome

• Many reporting packages can report on this.

  For instance, Awstats says

  AWStats is able to detect robots visits.[...] "Robots visitors" are reported separately in the "Robots/Spiders visitors" chart.

  EDIT: AWStats is free, written in Perl, and works just fine on Windows and with IIS logs.

  Pure.Krome : I've updated my original post. Bascially, i can't use PHP. has to be .NET based ... and preferably free.

  WerkkreW : AWStats is based on Perl, not PHP.

  Jacco : AWstats has only very basic bot-detection. Anything that has a non-robot useragent and does not request robots.txt is not recognised as a robot.

  Pure.Krome : re: Php/Perl. opps. I ment Perl, but same thing.. i don't want to install extra stuff besides the app iteself.
  From Paul

• Here is a pretty good list of known bot user-agents. I am not familiar with any free web log analytic tools written in .NET, but you could easily parse the logs yourself if all you are looking for are bots.

  http://www.user-agents.org

  From WerkkreW

• I use WebLog Expert when I want to examine our logs. There is a free version (lite) but that doesn't include the scheduler or command line functionality. The software itself runs as a desktop app (either remotely with ability to FTP logs or can be installed on the server), and can output reports in CSV, PDF and HTML.

  From Dan

• Hello,

  why not outsourcing your web stats ? With solution like sitemeter, you just put an extra javascript, and they will give you stats on your site.

  Web stats may be a pain for admin, doesn't always worth it to implement in house solution

  From Mathieu Chateau

• Google Analytics anyone? Free, Javascript based, so no need for server side languages. In depth coverage of hits, bots too.

  Pure.Krome : I'm already using Google Analytics. It has stats on it's own robot hits?

  Gr33n3gg : Hm, sorry I misled you, I thought GA did bots too. The bots/crawlers can't execute Javascript, Awstats would work for you since it reads the server logs.

  Pure.Krome : I didn't know this, but Google Analytics actually records how many times it hits your site. That said, it doesn't count the OTHER bots, though.
  From Gr33n3gg

Cannot connect to Hamachi today

I can't connect to Hamachi today, anyone has the same problem?

The hamachi connection before today was OK

From serverfault Ngu Soon Hui

• This is not really the appropriate forum for such a question.

  See: http://community.logmein.com/logmein/board?board.id=11

  From Justin Scott

Elegantly Enforcing SSL in IIS 6.0

I have a website running under IIS which has an SSL certificate applied. We would like to enforce HTTPS usage for the website which is easily done by checking the "Require Secure Channel" box, but this will immediately break the ability for people to connect over HTTP (as designed).

What I'd like to do is find a way to automatically redirect people from HTTP -> HTTPS if they type in the wrong thing (or connect from an old bookmark).

Is there a way to do this without creating a second website in IIS?

From serverfault kdmurray

• If you need to do it at the server level then you will have to create another site and have it forward the the https site.

  If you can edit the code, you can not enforce ssl at the server level, instead you can do it in your website by detecting if the url starts with http: and redirecting to the same url with https: instead.

  Ryan Bolger : Editing the code won't help since the request will never make it to the code if the https is enforced at the IIS level.

  sysadmin1138 : However, if the flag is turned off (allow http), their code WILL run and redirect cleanly.

  Jimmie R. Houts : @Ryan, sysadmin1138 is correct, I have updated my answer to be more clear.

  kdmurray : OK. Thanks. I pretty much figured that was the answer. I'd really love to see an "auto-forward" feature in IIS 8 to make this sort of thing automatic. Seems a waste to clog up the IIS metabase with extra sites that are simply redirects.

  kdmurray : As for the code option, the HTTP variant of the site needs to be turned off, that's the whole reason for needing to enforce SSL to begin with. Simply doing a redirect in code (unless it's done at a high level so that it happens on any page) could potentially allow for bookmarked pages to be viewed over the insecure link.
  From Jimmie R. Houts

• I posted this on StackOverflow

  If the server/site/vdir is configured using the "Require Secure Channel" setting, the response from the server will be a "403.4 Forbidden: SSL is required to view this resource." error or a "403.5 Forbidden: SSL 128 is required to view this resource.".

  You can actually customize the 403.4 or 403.5 error to redirect back to HTTPS. Create a VDIR under your site with NO SSL Requirement (**This is Important) - I use "CustomError". Create an ASP File inside this directory called 403_4_Error.asp containing the following:

  <%@ LANGUAGE="VBScript" %> 
  <%
  if Request.ServerVariables("HTTPS") <> "on" then
      sServer = Request.ServerVariables("SERVER_NAME")
      sScript = Request.ServerVariables("SCRIPT_NAME")
      sQuery  = Request.ServerVariables("QUERY_STRING")
      Response.Write("https://" & sServer & sScript & "?" & sQuery)
  end if
  %>
  

  Edit the server/site/vdir's Custom Error property for 403.4/403.5 and set the MessageType to URL and the URL to "/CustomError/403_4_Error.asp".

  Note that ASP is used, you could easily use ASP.net or any other scripting language.

  sysadmin1138 : This is the method we use.

  kdmurray : Thanks for that. I really was looking for a server-side (non-code specific) solution for this as we're marking the entire domain as requiring SSL. I'll keep this in mind for any future apps.

  Christopher_G_Lewis : This is completely server side - it executes in the server's error page handler.

  Jimmie R. Houts : +1 never thought of using the error page to do this. Very elegant.

  Whisk : +1 for this, it's extra easy to do in IIS7 as you can specify a redirect straight from the error customisation page

  lextm : Actually this trick is also documented in KB 839357. http://support.microsoft.com/kb/839357

  Christopher_G_Lewis : Yes, I believe I started with that article - Exchange was the initial premise for our needs. The above script will work with any site, not just Exchange OWA.
  From Christopher_G_Lewis

• We ended up doing this in our application code; our page base class checks early on whether it's in HTTP, and redirects to HTTPS as required. It really depends on how much you trust your application code :)

  From $pageUsers[$post.author].name

How can I tell what version of IIS is installed?

I've got a box running WinXP professional and some directions from Microsoft KB about SSL certificates, for IIS 5.0 and 6.0. How can I tell which version of IIS is currently installed?

Edit: The OS is Win2k3, not XP.

Sorry for the confusion,
Management

From serverfault jldugger

• Windows XP has IIS 5.1 installed, so use the IIS 5.0 procedure. See this article for an overview of IIS 5.1:

  A look at IIS 5.1 in XP Pro - What's different from IIS 5?

  IIS 5.1 is a feature only to be found on Microsoft's XP Pro operating system. It is not installable (reliably) on XP Home. Additionally, there are no plans to update IIS 5.0 on Windows 2000 to IIS 5.1.

  IIS 5.1 is basically the same engine as IIS 5.0 but since XP is a client operating system, it has the built in limits that are customary for Microsoft’s client operating systems—such as connection limits and only one Web site. Even though based on IIS 5, there are significant differences from IIS 5.0 that you should know about.

  You could also open a page in the browser which runs this simple asp Script:

  <% 
      response.write(Request.ServerVariables("SERVER_SOFTWARE")) 
  %>
  

  ---

  As a side note: it is interesting that IIS (beginning with Windows Server 2000/version 5.0) can't be upgraded without upgrading the operating system. Every Windows version has it's own IIS version:

  Windows Server 2000      5.0
  Windows XP Professional  5.1
  Windows Server 2003      6.0
  Windows Vista            7.0
  Windows Server 2008      7.0
  

  jldugger : "only one Web site". Does that include Microsoft SharePoint Administration? Because if not, then there's clearly two websites running.

  splattne : Try to add more that to web site (not virtual directory or web apps under the default web in IIS 5.1 on Windows XP). You can't do that.
  From splattne

• You can look at %SYSTEMROOT%\system32\inetsrv\inetinfo.exe. Right-click and get properties, click the Version tab.

  Edit: All kinds of fun :)

  You can also look at an HTTP response header

  1. telnet mywebserver 80
  2. type in HEAD / HTTP/1.0 [enter][enter]
  3. Look at the line that starts with Server:
  jldugger : Interesting; it claims version 6.0.

  squillman : On XP Pro?? Never seen that... Mines says 5.1.

  splattne : @jldugger: I doubt it. I tried to telnet 80 a XP machine. It's 5.1. You must have been connected to a Windows 2003 server. (typo in server name?)

  jldugger : We're all kinds of crazy around here. A few comments about your edit: 1. I think there's an unallowed space between / and HTTP 2. It just returns the normal HTML webpage; were you expecting something different?

  jldugger : Ok so it turns out I'm stupid and remote desktop is confusing when full screened. I noticed the WinXP Pro from my own box and believed it was from the remote host. It's win2k3.

  squillman : Oops. Turns out I'm stupid too and used angle brackets in my original edit around the enters...

  squillman : ANNND, make it HEAD instead of GET so you don't get the whole stinkin' home page, just the headers.

  squillman : And, no. You want the whitespace between the / and HTTP. The / says to get the root of the site. Thus the change to HEAD... (sorry for the comment spam)
  From squillman

• As a more general answer, not specifically aimed at your question, Microsoft has a support article which lists all of the versions and the operating systems that provide each one.

  squillman : Yah, that was the page I was looking for! :) +1
  From Gavin McTaggart

Diffing two filesystems

Is there any sensible way to diff two filesystems?

On occassion, you have a box that is working, and, say, perform an update and it no longer works. You have dozens of other machines that are (theoretically) identical, save for user data [and information tied to the specific hardware, and, say, ssh keys and so on]. Is there any way to determine what is different between the failed machine and one of the good machines?

I've tried mounting volumes and running ls -l or md5deep and recording the results and diffing that. Heck, I've even tried mounting two filesystems and just plain diffing them (with a graphical tool, mind you). The results have not been as useful as I'd expect.

Part of me wonders about having a database of files on each machine and what their md5sums are, with a cron job to update it on occasion, but I'm not even sure that would be as useful in practice as it sounds in theory.

So, is there any sensible way to diff two filesystems (or even, diff one filesystem against itself at a different time)?

From serverfault Clinton Blackmore

• Use rsync with --dry-run option.

  For monitoring local changes, use tripwire.

  Clinton Blackmore : Now that's brilliant. Thanks.

  Zoredache : Also check out AIDE as the OSS solution (http://www.cs.tut.fi/~rammer/aide.html)

  Ian Kelling : For those of us who use this regularly, -n = --dry-run

  Ian Kelling : If you want to be thorough, use the checksum option of rsync.

  hendry : Rsync does not give any intelligble output. hendry@x61 tmp$ rsync -cPnra foo bar sending incremental file list foo sent 64 bytes received 15 bytes 158.00 bytes/sec total size is 0 speedup is 0.00 (DRY RUN) Is bar different to foo, vice-versa or the same ?
  From vartec

• There is a tool called fsdiff [I could not find the homepage] that will compare a filesystem to its transcript. A free, multi-platform tool called radmind includes it and claims,

  At its core, radmind operates as a tripwire. It is able to detect changes to any managed filesystem object, e.g. files, directories, links, etc. However, radmind goes further than just integrity checking: once a change is detected, radmind can optionally reverse the change.

  From Clinton Blackmore

how to install fuzzystrmatch.sql pg_trgm.sql on postgresql on mac os x

I have installed postgresql according to this, but I also need fuzzy string match functions, but I have no idea how to install these.

From serverfault Alex. S.

• Usually with a sql file full of functions, it is something like this:

  cat fuzzystrmatch.sql | psql {database name}
  

  Alex. S. : the problem is, where can I get that file?, it isn't in my system.

  Milen A. Radev : Ask the guys from "PostgreSQL for Mac".
  From jedberg

• For the Mac, the .sql files are in a folder such as /Library/PostgreSQL/8.3/share/postgresql/contrib

  Clinton Blackmore : For me, it was in /Library/PostgresPlus/8.3/share/contrib .
  From Frank J

Windows 2008 Corrupt TCP/IP? (Ping access denied)

I'm having a weird problem on two Windows 2008 VMs (inside HyperV) running NLB. They were working at one point, but now TCP/IP appears to be corrupt. I cannot ping anything if logged in with my Domain Admin account:

>ping 10.1.1.1
Unable to contact IP driver, error code 5,

I tried resetting winsock and ip with netsh. That makes no difference:

netsh int ip reset
Resetting Echo Request, failed.
Access is denied.

I uninstalled NLB, I removed the synthetic NIC and added a legacy NIC from Hyper-V -- no dice.

The weird thing is that it works if I login with the built-in Administrator account, but not my Domain Admins account. UAC is disabled, so this should just work.

Any ideas apart before I call PSS?

Edit: Can't ping anything if I don't login with the built-in Administrator account. (UAC on or off makes no different.) Can't ping 127.0.0.1 (access denied). NICs are the HyperV ones.

I installed SP2 for Windows 2008, no change.

From serverfault MichaelGG

• I've run into a few things that still need you to 'Run as administrator' on Vista and Win2008. Even if UAC is disabled. Not sure why, but try it and see if it helps out.

  -JFV

  MichaelGG : Tried that too - no luck.

  JFV : The only other thing I can recommend is to update the drivers for your network card and see if that helps.
  From JFV

• Can you ping localhost? If so then TCP/IP should be fine. Perhaps it's got a domain firewall setting enabled that it shouldn't have?

  MichaelGG : Nope. Can't ping anything - get the access denied error when trying to ping. Disabling the firewall doesn't seem to change anything.

  mrdenny : Not even local host? That's surprising. I'd give Corey's regkeys a try then.
  From mrdenny

• I ran into this same problem on Server 2008, completely out of the blue. Tried the same steps as you also, with no luck. I ended up dumping the Winsock and Winsock2 settings from the registry from another (working) box and using those.

  Download from:

  coreygilmore.com/uploads/2009/05/winsock-reg.txt

  and

  coreygilmore.com/uploads/2009/05/winsock2-reg.txt

  if you want to give it a shot, just back up HKLM\SYSTEM\CurrentControlSet\Services\WinSock and HKLM\SYSTEM\CurrentControlSet\Services\WinSock2 first to be safe. After you back those keys up, delete them from the registry, then import and reboot. Dumped from Server 2008 x86.

  MichaelGG : That's a good idea. Thanks!

  Corey Gilmore : I should mention that my problem was also with virtualized Server 2008 (running on VMware Workstation 6.5). I had three Server 2008 VMs on a new forest/domain; Domain Controller, Exchange 2007 server, BES. My user account (Domain Admin) and the BESAdmin account could *not* ping on the BES, but the local Administrator could. Both BESAdmin and my account could run ping on other boxes. The referenced error message is not a firewall issue and it's not a permission issue as other users are suggesting. In my case the box responded to pings too, the problem was weird Winsock corruption.
  From Corey Gilmore

• Check the firewall is not blocking PING (ICMP - listed under file and printer sharing).

  I had a similar problem previously where the host and virtual machine both had the firewall enabled and both were blocking PING, caused all kinds of havoc.

  MichaelGG : Even tried turning off the firewall - no luck.
  From Robert MacLean

• If your Firewall is enabled...you should enable File and Printer sharing.....File and printer sharing allow ICMP packets which is being used by PING

  OR

  If you dont need Windows Firewall...make sure to turn it all off....3 levels...

  From Eric

XP Enlarge when in high resolution?

Someone I know has bought a new LCD monitor, with a native resolution of 1680 * 1050

However, they can't read stuff at that resolution because its too small for them. They need to use something a fair bit lower - 960 * 600 or one above.

However, that being non native, its kind of blurry/grainy - the screen isn't doing the downscale/interperlation very well.

To my mind, it seems plausible that there might be some software available that can let the monitor run at its native resolution, but enlarge everything to the same size as a lower resolution - it must be an increasingly common problem?

What solutions is anyone here aware of?

From serverfault benlumley

• You can experiment with changing the DPI setting in windows (it's a tab in the same place where you select your driver and such in windows). Depending on how much you need to make the display fonts bigger, this may work for you.

  Oskar Duveborn : This is the way to do it right now, be prepared that not all applications are aware of this and may look or work wierd...

  benlumley : Yeah - DPI doesn't seem to affect everything, so its not that useful.
  From Roy Rico

• You can use the screen magnifier in the accessibility options.

  benlumley : yeah, that wasn't well received as it seems to only do part of the screen at once. A kind of permanent zoom is what is wanted.
  From WerkkreW

• Maybe try "Left Alt"+"Left Shif" and PrtScn. Then see if you can adapt the colours :-)

  Also look at "Display Propeties"+"Appearence" - font sizes can be changed there.

  From Tubs

Shutting down a Linux based NAS remotely and reliably

I have a Linux based (Debian 5.0) NAS that I need to shutdown reliably. When ssh'd into the NAS box I issue the following command:

sudo shutdown -h now

I've also tried 'poweroff' and su'ing into root and running the commands with the same results. It shuts down all the processes and everything fine but does not always 'halt' the system. I need it to power down into the S3 state so that I can reliably use Wake On LAN to start the unit as needed. For what it's worth, wol works great and works every time as long as the unit is in it's S3 state.

From serverfault

• Try the halt -p or poweroff commands.

  From Justin Scott

• Humm could be a number of things. First thing I would try is to pass either

  acpi=force
  

  or even

  acpi=off
  

  to the kernel command line to see either of them works.

  Next I would try unloading the wireless drivers before you shutdown (if possible).

  From Nathan

• I have had similar problems, I just issue a halt instead, and it seems to work fine.

  From Brad Gilbert

• I finally diagnosed the problem with this particular NAS backup server. It appears that the graphics card died at some point. Since this unit is run remotely via ssh it's only two lifelines are it's ethernet cable and it's power cord. Consequently, the dead graphics card went totally unnoticed. Putting in a new graphics card has solved the problem completely. I also did a lot of work to the unit to make sure that it has better airflow and much less heat build up since it is located in a place that is not very well ventilated and I believe that this may have been the cause of the graphics card failure. It could have just been the age of the graphics card but since I had the unit out I decicded to update/upgrade it's thermal properties just to be on the safe side.

  Matt Simmons : weird, but glad it works!
  From

How to get apache2 to redirect to a subdirectory

I am running apache2 on debian etch, with multiple virtual hosts.

I want to redirect so that http://git.example.com goes to http://git.example.com/git/

Should be really simple, but google isn't quite cutting it. I've tried the Redirect and Rewrite stuff and they don't quite seem to do what I want ...

From serverfault Hamish Downer

• Feel a bit silly - a bit more googling turned up the answer I was after:

  RedirectMatch ^/$ /git/
  

  Basically redirecting the root, and only the root.

  From Hamish Downer

• On my own website, I used a 301 redirect by placing the following in index.php in the root directory:

  <?php
  // Permanent redirection
  header("HTTP/1.1 301 Moved Permanently");
  header("Location: http://git.example.com/git/");
  exit();
  ?>
  

  David Zaslavsky : That works, but using PHP just to do a redirect is overkill (and not as efficient as Apache directives).
  From Joseph Sturtevant

• You've got the correct answer there with the redirect. You have to be careful when redirecting everything to somewhere else, since you can get recursive redirects there. This happens if you want to put up a maintenance page.

  From Rory McCann

• You can use Redirect directive.

  <Directory />
     Redirect permanent / http://git.example.com/git/
     ...
  </Directory>
  

  From $pageUsers[$post.author].name