|
Messages are lost because Outlook Express' database system isn't transactional. It doesn't guarantee that the data has actually been written to the file. The NTFS is transactional only for file system metadata - it doesn't protect the data itself, it only ensures that you don't lose the whole volume.
Basically, I think you hit the small window of time between OE asking the OS to write to the file and the OS actually doing so. This write caching is perfectly normal but you do have the possibility of losing data. However, OE has already told your mail server to delete the message from its store.
I'm afraid it's gone. You aren't going to see that message again.
In a corporate environment, it's usually much better to use Exchange or another enterprise messaging server. Exchange understands SMTP, the messaging protocol used to send messages between servers. Exchange's database is transactional, and it won't report successful delivery to the sending server until it has at least committed its transaction log to the disk, containing the new message(s). That means, if the power fails, either the message has been recorded and Exchange will roll it forward into the main database, or it hasn't, in which case the sender still has the message and will retry sending it later.
Finally, it's a lot easier to protect the server against power failures than it is to protect all the workstations. Even a small UPS (uninterruptible power supply) can buy enough time for the server to shut down gracefully.
Stability. What an interesting concept. -- Chris Maunder
|
|
|
|
|
Thank you very much, your give a good idea to deal with this problem. Roger (the person who participate to answer this question) also give me a good idea too. Before that I really cannot figure out what is the main problem. Thank for a greate idea 
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Mike Dimmick wrote:
it's a lot easier to protect the server against power failures than it is to protect all the workstations.
Excellent observation. A good UPS is a tiny investment to protect priceless resources, and the server is the smart place to invest first.
"My kid was Inmate of the Month at Adobe Mountain Juvenile Corrections Center" - Bumper Sticker in Bullhead City
|
|
|
|
|
My best guess is that, because the shutdown was not an orderly one, Outlook Express probably had not yet saved the message to disk. I don't know how often OE saves emails, but it's possible that it only does so as part of the shutdown process, or at programmed intervals that hadn't occurred between the time you received the message and the power went out. In either case, the email would have existed only in RAM, and it will not be possible to recover it.
"My kid was Inmate of the Month at Adobe Mountain Juvenile Corrections Center" - Bumper Sticker in Bullhead City
|
|
|
|
|
Hi, I have a problem when I change my harddisk to another computer. My harddisk contain winme, because it have a problem with mainboard I have took it out. I take to another computer and remove the harddisk of another computer and put it in. After that I go to CMOS (when it first start) and detect for the new harddisk. During detection, both harddisk and CDROM is found but when I quit CMOS (quit and save) and start windows, there is no CDROM drive appear in windows explorer. I try to check the registry, clean virus and check some hardware configuration but there is no way to show it in windows explorer. Does anybody have idea or solution with this problem?
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Dear anyone who can help,
I am a programmer currently working for an educational software company. We have a windows product that uses a centralized database on a server that serves entire school systems with data over a WAN. Right now it is in the beta stage of development.
We are having a delema and are hoping somone has the answer. The problem is: each of the workstations, and there can be thousands, need to be updated with patches every so often as they become availible. We can write an app. that runs when the workstation program starts, to check with the server to see if there is a patch availible and wheather or not the latest one has been applied or not. Then download it and install it. The only catch is, most times the workstation app. is started by students, and they are often times logged on to windows as a limited users. And when you are logged on to windows as a limited user, windows will not let you save files, except to certain, rather useless places for our purpose. Also, windows will not let you run programs that update or install things.
Basically what we want is for the app. to start up and automatically update itself if there are updates availible. Thus making it much easier and less of a hassle to computer illiterate teachers, administrators and students. Who would otherwize have to apply a patch to each and every single workstation. That way the only person who has to worry about it is the top level sys-admin and all he/she has to do is post the latest avalible updates on the centralized server.
Is there a way to achive this?
Thanks for your help, If you require more info just ask.
|
|
|
|
|
Rewrite it as an ASP.NET application? No update problems there!! This is a large part of the reason that web applications are so popular.
If you are needing to patch COM objects, i.e. written in VB6 or C++ using COM, then you have very limited options. There are some products out there that let an administrator PUSH updates to the client, such as Microsofts SMS application.
If you have a .NET application, you have more options, because .NET apps support XCOPY deployment, i.e. you can just copy the program to a location that the user has rights to. There are also some prebuilt .NET components that help automate the process.
my blog
|
|
|
|
|
Before you do anything, are you sure you need this at all? Just because version 5.1 is current doesn't mean versions 1.0-5.0 needs to stop functioning. If the user doesn't need new features and the bugs haven't hurt them (yet)there is no reason they have to update. Tell them they must upgrade to get support, but otherwise let them fall behind.
In the case of schools it is normally best to update between terms so that students don't have to learn the next version while putting finishing touches on the night before it is due. (and a lot of students do this even if it is a bad idea) Just tell the teachers what the bugs are, and have them not assign something that would hit those bugs.
You could write a program that runs (service? something else?) once a day, checks for updates, and if they exist loads them.
I'd recomend checking out bit-torrent (or whatever it is called), or something like that, otherwise when updates do happen your servers will be swamped. (perhaps with something to prefer local subnets, since you can expect several machines to be on the same subnet in most cases)
Make sure that you carefully test your application before each release. It would be a very bad thing for it to update itself, and then fail to work because of some obscure bug that only happens on strange systems... Perhaps you want to consider a back out program so if your update does break for someone you can walk the teachers through restoring the old one on the phone. (and check this ability twice as much)
Oh, and make sure auto-updates are easially disabled for those with good admins who will do this themselves.
|
|
|
|
|
Hi All - I'm not much of a SMTP or Exchange guru, so I thought I'd throw this out:
I have an app that needs to send to an SMTP server (within IIS) to send email. What I'm attempting to find out is if it is possible for any email that is sent to that SMTP server to:
1. Best case - be actually processed by an exchange server I specify.
2. Next best - be automatically forwarded to another email address that I specify.
Does smart host have anything to do with this???
Thanks!
Marcus Spitzmiller
"Why must life be so hard? Why must I fail at every attempt at masonry?" - Homer
|
|
|
|
|
Hi,
My current Project has the following Requirements:
1. Backup the entire contents of the hard disk including Operating System as a bootable Image.
2. In case of System Failure, user should be able to perform a full system-recovery operation using the Backed up Image of the Hard Disk.
I am new to this concept and request any sort of help on how to proceed with this.
Any help will be greatly appreciated..
Regards
-Ram
|
|
|
|
|
I am able to use WMI to connect to another computer and retrieve its printer's info (such as name, model, resolution, shared properties,etc). Now if that printer is being shared on the network, is there a way for me to retrieve all the computers/users that are using this shared printer?
|
|
|
|
|
Hi, I have a problem when load the user profile from the server. My purpose is store the user profile on the server and when the user go to log on an client, they will load the profile on the pc client that they log on. For example I have 10 computer, and user log on to the computer number 3 and change his/her desktop color to green. When shee log off and go to log computer number 8 (or the other computer), her profile will load from the server and change the desktop color to green (with the location of the my documents folder and the shortcut on her desktop. But  I tried many according to the books on creating the profile for the user and change the type of the profile, it still do not work. It display a an error message since the time that I log on to the network. Does anybody know that problem? If you have another idea please write a step for me to load the user profile from the server from any client machine.
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Hi ALL,
I had created an application in MFC environment for windows XP.
When I try to ran the same application in wind00 2000, the following problem occurs.
The problem with Windows2000 occurs when you try to open a browse dialog box to save or open a file. So file>open hangs the GUI, File > save as hangs the GUI.
Note:This is what I had edited in stdafx.h code.
The CFileDialog class encapsulates the Windows common file dialog box, which provides an easy way to implement the standard File Open and File Save As dialog boxes."
In order to access Win2000 specific APIs etc, one needs the following #define in the application's stdafx.h (before any other #includes)
#define _WIN32_WINNT 0x0500 -(1)
The work-around for this is to #undef _WIN32_WINNT before the #include <afxext.h> in the application's stdafx.h: #define _WIN32_WINNT 0x0500 // allow Win2000 specific calls (2a)
#define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers (2b)
#include <afxwin.h> // MFC core and standard components (2c)
#undef _WIN32_WINNT // allow CFileDialog to build with the correct size (2d)
Ref:http://www.google.co.in/search?q=cache:zjjzsWFj4ygJ:www.thecodeproject.com/w2k/access_api.asp+cfiledialog+win2000+memory&hl=en
After adding the above piece of code also the problem remains the same.
With Best Regards
Lavate Mallikarjun Avanna
|
|
|
|
|
Hi ALL,
I had created an application in MFC environment for windows XP.
When I try to ran the same application in wind00 2000, the following problem occurs.
The problem with Windows2000 occurs when you try to open a browse dialog box to save or open a file. So file>open hangs the GUI, File > save as hangs the GUI.
Note:This is what I had edited in stdafx.h code.
The CFileDialog class encapsulates the Windows common file dialog box, which provides an easy way to implement the standard File Open and File Save As dialog boxes."
In order to access Win2000 specific APIs etc, one needs the following #define in the application's stdafx.h (before any other #includes)
#define _WIN32_WINNT 0x0500 -(1)
The work-around for this is to #undef _WIN32_WINNT before the #include <afxext.h> in the application's stdafx.h: #define _WIN32_WINNT 0x0500 // allow Win2000 specific calls (2a)
#define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers (2b)
#include <afxwin.h> // MFC core and standard components (2c)
#undef _WIN32_WINNT // allow CFileDialog to build with the correct size (2d)
Ref:http://www.google.co.in/search?q=cache:zjjzsWFj4ygJ:www.thecodeproject.com/w2k/access_api.asp+cfiledialog+win2000+memory&hl=en
After adding the above piece of code also the problem remains the same.
With Best Regards
Lavate Mallikarjun Avanna
|
|
|
|
|
Hey all. I've been stupid enought to take on to install Exchange 2003, so I suppose I'll just have to blame myself but I'd still appreciate some help on this:
When I try to enable SSL for OWA in [Exchange] System manager, the textbox for filling in the SSL port is grayed out, so I have to enable SSL from IISAdmin instead and it works. The problem is that when the server is restarted the settings I do in IISAdmin is overwritten with the ones from System Manager, so SSL dissapears
![Rose | [Rose]](https://codeproject.global.ssl.fastly.net/script/Forums/Images/rose.gif) To anyone with a decent (scripted?) solution or suggestion.
(I Have a CA on the same machine.)
|
|
|
|
|
Is this on IIS 5 or 6 (Windows 2000 Server or Windows Server 2003)?
Our Exchange Server 2003 installation on Windows Server 2003 is currently using a test certificate from Thawte. I can't find any options for enabling SSL in Exchange System Manager on this server - where are you looking?
If I try to create a new HTTP Virtual Server through ESM, and choose Advanced next to IP Addresses, that does offer an SSL port field. It's initially grayed out because you've not yet added a certificate to this new virtual server, I think.
Unless you want to expose a different set of mailboxes to users connecting to the server with different IP addresses or (HTTP 1.1) host headers (e.g. you're hosting two separate domains on the same server), I'd stick with enabling SSL on your default server.
In the Internet Information Services management console, expand Web Sites. Right-click Default Web Site and choose Properties. Go to the Directory Security tab. Click Server Certificate, then follow the wizard to add the certificate to this site.
Once you've done this, you can assign the SSL port to something other than the default using the Web Site tab. I don't recommend doing this because it's unexpected by the users, and doesn't offer much additional security.
If you want to ensure that OWA users can only connect using SSL, right-click the Exchange virtual directory under Default Web Site and go to the Directory Security tab. Here, click Edit under Secure Communications and check the Require Secure Channel option. Remember that SSL protects your users' passwords and data, but does not provide access control. Unless you open your Kerberos ports on the firewall, authentication data (username and password) is sent using HTTP Basic Authentication, i.e. in clear text, so I recommend requiring SSL.
Stability. What an interesting concept. -- Chris Maunder
|
|
|
|
|
Well, as I said, I do get SSL to work, but the SSL settings are deleted when the virtual directories are read from ESM when restarting the services, so I need to enter them there instead of in IISAdmin. I have several ips on the Exchange server to enable two companies to use the same one with different public folders and only port 443.
Thank you for your exchaustive answer though.
|
|
|
|
|
hi everyone 
i got one problem,i want to retrive all the informations like opened files,vistited webpages ect with specific date,which was cleared/deleted, by just using regidit, is that possible?bcos i think windows has this option.
curious 2 hear the ideas from u 
------------------------------
Lar Sopheak
|
|
|
|
|
Sopheak Lar wrote: hi everyone
Hello baby  ,
Sopheak Lar wrote: i got one problem,i want to retrive all the informations like opened files,vistited webpages ect with specific date,which was cleared/deleted, by just using regidit, is that possible?
No it isn't. Regedit just store an information of the specific folder which store the history file. It do not store the each link of open or visit webpage. Try to log on as any user account and navigate regedit to HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\User Shell Folders then you will find the path to each specific folder for the current user account. The folder that store the recent file that might offer your needs:
- %UserProfile%\Recent<br />
- %UserProfile%\Local Settings\Temporary Internet Files<br />
- %UserProfile%\Local Settings\History
Sopheak Lar wrote: curious 2 hear the ideas from u
Curious to meet you but you never want to meet me 
|
|
|
|
|
Hi, i have a question about set the permission for folder in MS-DOS mode. It mean that you stand on MS-DOS prompt then try to set the permission for folder by using the command line. Is it possible or not?
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Try cacls /? 
"My kid was Inmate of the Month at Adobe Mountain Juvenile Corrections Center" - Bumper Sticker in Bullhead City
|
|
|
|
|
Hi, I have a problem when I send or receive an e-mail message. There is a webpage pop up and show me that my antivirus is expire. Do you know why it have this problem?
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Suppose i have a Server or a domain controller and lots of clients (98,ME,XP..)
and i want to install a piece of software on all of the clients
is there any way by which i can automate this installation
from the Server and go physically to every client and install
Thanks
C makes it easy to shoot yourself in the foot; C++ makes it harder, but when you do, it blows away your whole leg
http://www.mastishk.com
|
|
|
|
|
In my opinion, it might be depend on the type of the software that you install. If you have a client/server networking and you install norton antivirus corporate edition 8.0 it will contain an option for you to choose the computer name (client name) from the network and it will proceed the installation process for each client. For the tool or utilities you need to use remote desktop to access the client or use the terminal service to access the remote client. This just my opinion, if the other people find a mistake in this paragraph please correct it and give the right one.
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
Hi, I have a problem when I share a resource on the network. My computer use windows 2000 professional. I share a folder that contain a picture call "rt04" set the full permission to Everyone user. When I access from windows xp client it prompt me for user name and password. I supply "administrator" as the user name and the password of adminstrator account on my computer then it could access my share files. But when I start to access the file from windows 98 it display a dialog box with only one textbox for password. The "Resource:" labeled stand for "\\[ComputerName]\IPC$" and a check box labeld "Save this password in your password list". From here I type the password of the administrator account on my computer but it is not work, it just display the dialog box again and said that the password is incorrect. I have two question in here:
1. Why I could not access the share file from on my computer from the computer running windows 98 (windows xp is working fine  )
2. The check box labeled "Save this password in your password list". What does the password list mean, where is the list that it store the password in my computer.
Please help me 
Roath Kanel
APO-CEDC
Save Children Norway-Cambodia Office
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
