What are "transactional" file operations? - winapi

I was browsing the Win32 API functions for file and directory management operations. I saw that some of those functions has their so called "transactional" counterparts.
Examples:
CreateDirectory and CreateDirectoryTransacted
RemoveDirectory and RemoveDirectoryTransacted
CreateFile and CreateFileTransacted
CopyFile and CopyFileTransacted
I read explanations of these transacted functions, the Wikipedia article Transactional NTFS and this MSDN Magazine page. But because of the heavy terminology (for me) in these pages, I didn't clearly understand these explanations. They all come to a common consensus that these functions are "atomic". But as far as I understand from the word "atom", it is a nucleus with spinning electrons around it...
Can you please explain me in basic and simple English sentences, what are the purposes and operations of these functions? Why and when would one prefer transacted version of an API function?

Why and when would one prefer transacted version of an API function?
There are a couple of scenarios given in the link I quoted above.
One of these is the use case of an installer application, which needs to copy/install several files to different locations and then maybe perform some updates to the registry. Before the installer runs the system can be considered consistent. Once the installer has done all its work, the software is completely installed and the system is again in a consistent state.
If, however, the computer crashes during the installation process it may not be trivial to determine which steps of the installation procedure have already been performed successfully before the crash and which have not. Transactional operations can give support in this situation by 'automatically' restoring the consistent system state as it was before the installer ran, if for any reason the installation fails in the process.
As Microsoft states, the transactional file system operations have never been adopted much by developers, which may serve as an indication that the functionality is not really needed for the vast majority of applications, or, that there are easier ways to achieve the desired result in an application-specific way, for which MS gives examples too.
Besides, the concept of "atomic" operations is present in different areas of software development, for instance in concurrent programming or in database management systems. See the Wikipedia article.

In short, a transaction (be it a file system, database or bank) will only be completed if no errors occurred in the process.
Using a non-transactional file system and API, say you have a file containing:
AAAA
Now you want to fill the file with all B's, but while doing so in the middle the power is lost and not all data is committed to the disk. Now you have an inconsistent state when you read the file back (after power returns):
BBAA
Remember FAT and scandisk?
Now with transactions, the file system basically first write the changes to a different location on the disk, and change the "file data location pointers" inodes towards the new location of the data only when finished, marking the space the old data occupied as 'available' again.
You don't need Transactional NTFS (TxF) for this, as 'standard' NTFS also promises to ensure consistency:
NTFS is a recoverable file system that guarantees the consistency of the volume by using standard transaction logging and recovery techniques. In the event of a system failure, NTFS runs a recovery procedure that accesses information stored in a transaction log file. The NTFS recovery procedure guarantees that the volume is restored to a consistent state. Transaction logging requires very little overhead.

IMPORTANT: Note that Microsoft marked the entire "Transactional NTFS" API as deprecated and strongly discourages its usage.
See Alternatives to using Transactional NTFS.

Related

How to detect Windows file closures locally and on network drives

I'm working on a Win32 based document management system that employs an automatic check in/check out model. The model it currently uses for tracking documents in use (monitoring the processes of the applications that open the documents) is not particularly robust so I'm researching alternatives.
Check outs are easy as the DocMgt application is responsible for launching the other application (Word, Adobe, Notepad etc) and passing it the document.
It's the automatic check-in requirement that is more difficult. When the user closes the document in Word/Adobe/Notepad ideally the DocMgt system would be automatically notified so it can perform an automatic check in of the updated document.
To complicate things further the document is likely to be stored on a network drive not a local drive.
Anyone got any tips on API calls, techniques or architectures to support this sort of functionality?
I'm not expecting a magic 3 line solution, the research I've done so far leads me to believe that this is far from a trivial problem and will require some significant work to implement. I'm interested in all suggestions whether they're for a full or part solution.
What you describe is a common task. It is perfectly doable, though not without its share of hassle. Here I assume that the files are closed on the computer where your code can run (even if the files are stored on the mounted network share).
There exist two approaches to controlling the files when they are used: the filter and the virtual filesystem.
The filter sits in the middle, between the process and the filesystem (any filesystem, either local, network or fully virtual) and intercepts file requests that go to this filesystem. Here it is required that the filter code is run on the computer, via which the requests are passed (this requirement seems to be met in your scenario).
The virtual filesystem is an endpoint for the requests that come from the applications. When you implement the virtual filesystem, you handle all requests, so you always fully control the lifetime of the files. As the filesystem is virtual, you are free to keep the files anywhere including the real disk (local or network) or even in the cloud.
The benefit of the filter approach is that you can control individual files that reside on the real disks, while the virtual filesystem can be mounted only to the new drive letter or into the empty directory on the NTFS drive, which is not always fisible. At the same time, sitting in the middle, the filter is to some extent more restricted at what it can do, and the files can be altered while the filter is not running. Finally, filters are more complicated and potentially error prone, as they sit in the middle and must play nice with other filters and with endpoints.
I don't have specific recommendations, but if the separate drive letter is an option, I would recommend the virtual filesystem.
Our company developed (and continues to maintain for the new owner) two products, CBFS Filter and CBFS Connect, which let you create a filter and a virtual filesystem respectively, all in the user mode. Those products are used in many software titles, including some Document Management Systems (which is close to what you do). You will find both products on their website.

Realtime one-way mirroring of a SQLite database

I am dealing with a 3rd party application that's running a SQLite 3 database with WAL (Write-Ahead Logging) on a local computer, and I'm looking to mirror that database (read only, this is a one-way mirroring) to another system. The challenge is that I'm running in a separate process, which seems to complicate things somewhat.
The database is being created and opened with a normal locking mode so there's no problem reading it from another process, but I'm trying to either find an existing implementation or get some pointers on where to get started. My understanding, based on other posts is that the standard sqlite update hooks (such as sqlite3_update_hook) will not work out of process.
A key issue is speed, I'd like to ideally be able to detect each update as soon as it happens and begin transmitting it. This means that most polling options would be out of the question, but even if they were, how would you detect the most recent changes?
I'm seeing two files that look promising: the actual WAL file (foo.db-wal), and that memory mapped index file (foo.db-shm). I'm hoping that those two contain the information I need to: A. Detect when changes occur in the database and B. Be able to grab just the incremental changes since the last update.
But a pointer to some existing solution would be much preferred... :-)
SymmetricDS might be the solution for you

How to implement a Definitive Media Library (ITIL DML)?

How to implement a Definitive Media Library (ITIL DML) ?
I would like to know some way to implement a DML based on ITIL.
Given a library of heterogeneous software the only solution that crosses my mind is to use a system file structure (with proper security and access permissions), however this seems very simple and if the library gets too big it will be hard to find software that search.
Is there any specific software for DML?
Many tools that offer CMDB management also offer DML management. Some options for this are ServiceNow and IBM's Change and Configuration Management Database.
If you are only looking for DML functionality, a binary repository manager, such as Sonatype Nexus or Artifactory, provides metadata tagging, version control, and many other useful features. Implementing a binary repository manager and proper procedures for maintaining it serves as an excellent DML solution.
There is nothing wrong with a file system to store software in form of completed, tested (software) configuration items which passed appropriate quality assurance test, etc. But because of the controlled it environment in ITIL, it is mandatory to establish an control structure for tracking the appropriate information of every CI or software in the DML. This records have to contain all relevant informations like version, build date, development, release date, etc. Only tested, confirmed and quality checked, and deployable CIs should be hold in this system.
Because of the extensive meta-data, some years ago, we created a DML depending on postgreSQL because the handling and management for the CIs along with the mandatory tracking (time when inserted, time and logging of access, access control, maybe licenses, etc.) and meta-data is much easier in a sql database. Of course, we had to build the structure for the meta-data into the sql-db, but that was not too complicated and for our straightforward DML it was sufficient. Building and managing the metadata with each ci was supposedly easier for our system than installing, configuring, learning and managing a whole third-party DML/CML system. A caveat is when we had to save whole system images from deployed, tested, already integrated and checked systems because their size were several hundreds of GB (With the newest DB-versions, this should now also be possible, the question is, if it is useful within a SQL-DB) . But we stored the disk-images on separate disks and tracked the meta-data in our tailored DML-system (our postgreSQL) along with the information where to access it.
The advantage was, that we could easily duplicate the DML and take it for example to customer, where we only had to set up or run our (postges-based) DML and were able to access all relevant CIs we needed to set up an heterogeneous network of the target system.
In other cases, maybe it is easier to rely on already existing third party solution, but the idea of a DML can be fulfilled with every storage system as long as the appropriate proecedures, meta-data, informations and access points to the overall life-cycle management are provided and maintained.
regards

multi-client inter-process communication on Windows, VB6

What is the best way for multiple client programs to
communicate with a single server program, all running
on a single Windows computer? All written in VB6.
I'd appreciate recommendations of how you might solve
this problem.
NOTE: we are working on transition to .NET, but have to
add a capability to the V6B version before the .NET will
be ready.
The possibilities include TPC connections, named pipes,
shared memory, messages, files, and more.
A client passes the server a string as input, and the server
combines it with data known only to the server, to generate
another string which is returned to the client. Both strings
are only about 100 characters long. The server is contacted
only when a new file needs to be opened, and so it is a very
low volume of communication... probably a flurry of 10 calls
within 15 seconds, followed by an hour of idle time.
But it is possible that two clients would choose about the
same time to request information. Blocking/Locking are certainly
acceptable, as the server will be done with each request in
well under a second, and several seconds of delay is unimportant
to any of the programs.
The server's algorithm is complex, and for several reasons important
to the application should not be replicated in each helper program.
That is the reason for needing a server.
Background:
I am adding capability to a large existing legacy program.
This single program has several other legacy programs which
act as helpers and are run when the user makes certain
choices. These programs are started with a shell command,
and are not just separate threads. For instance, one helper
loads new data from a DVD drive onto the hard drive. Another
helper just displays a chart of the current positions of
the planets.
This is a LARGE commercial legacy program that happens to be
written in VB6. We are working to convert it and all the
helper programs to .NET, but must first release a new version
under vb6 with this added capability. (Please don't tell me
to not use VB6, as we are already moving elsewhere.)
We need a temporary VB6 solution.
VB6 does TCP and UDP extremely well via the standard Winsock Control component included in Pro and Enterprise Editions. A lot of shadetree coders do seem to struggle with it though. This is probably the most obvious route since the only other native IPC in VB6 would be COM/DCOM and DDE, however MSMQ provided excellent support for VB6 as well.
The downside of IP-based protocols is their limited namespace and resulting high probability of collisions (64K port numbers, many set aside for standard applications, ephemeral port ranges, etc.). They're also somewhat "heavyweight" but considering the vast resources of even the oldest PCs still in service and your light traffic requirements you can ignore that in deciding.
Another option you've considered is Named Pipes.
This offers a number of advantages in your situation. For one thing the namespace is much larger requiring only a unique name, which in the post-Win9x era can be up to 256 characters long making uniqueness fairly easy to achieve. For another, as long as your firewalls permit "File and Print Sharing" you're all set on that front.
Also, for your application you only seem to require an RPC-style mechanism rather than arbitrary bidirectional streams or messages. TransactNamedPipe() calls in your clients might be ideal. Named Pipes work over a LAN, but within one PC they are quite fast and light weight.
While VB6 doesn't come with a Named Pipe component such a thing is fairly easy to create as long as extremely high performance isn't required. You can use Timer-based polling in the server instead of trying to implement overlapped I/O to get asynchronicity. I put one together a couple of years ago and have had good luck with this approach.
I published a fairly stable rendition of this a while back at PipeRPC - RPC Over Named Pipes. There is an older and a somewhat newer version there with examples of use and documentation. As designed, clients make "calls" passing a Byte array of request parameters and receiving back a Byte array of response results. You can also shove Unicode Strings though with no changes, letting the compiler coerce the types.
Just one "drop in" UserControl for both clients and servers.
Looking back at this question:
The server's algorithm is complex, and for several reasons important
to the application should not be replicated in each helper program.
That is the reason for needing a server.
If that's really the concern why not just create a shared DLL that all programs use?
For a one-off upgrade release to an existing VB6 application being moved to a newer platform, I would stress keeping the modification as simple and straightforward as possible. As a result, I wouldn't go down any routes involving shared memory or anything relatively unusual.
A few options, none perfectly simple, but at least some ideas:
Expose a COM object in the server code that performs the translation, and can be consumed by the client apps. The clients instantiate the object from the server as an out-of-process object, and let COM handle all the marshalling, etc.
Does the server have any network awareness? VB6 doesn't do sockets/tcp natively very well, but if you've had a reason to add that in, you might be able to leverage it to perform a socket-based connection and data exchange.
The server and client could each poll a common resource folder for the presence of a specific file that constituted inbound/outbound requests for the translation service you describe. Not very elegant, but it might be the simplest.
Just a few ideas to give you some things to think about. Hope that's helpful in some way. Good luck!

Windows Registry best practices

In what way is the Windows registry meant to be used? I know it's alright to store a small amount of user preferences, but is it considered bad practice to store all your users data there? I would think it would depend on the data set, so how about for small amounts of data, say, less than 2KB, in 100 or so different key/value pairs. Is this bad practice? Would a flat file or SQLite db be a better practice?
I'm going to take a contrarian view.
The registry is a fine place to put configuration data of all types. In general it is faster than most configuration files and more reliable (individual operations on the registry are transacted so if your app crashes during a write the registry isn't corrupted - in general that isn't the case with ini files).
Marcelo MD is totally right: Storing things like operation percentage complete in the registry (or any other non volitile storage) is a horrible idea. On the other hand storing data like the most recently used files is just fine - the registry was built for just that kind of problem.
A number of the other commenters on this post talking about the MRU list have discussed the problem of what happens when the MRU list gets out of sync due to application crashes. I'm wondering why storing the MRU list in a flat file in per-user storage is any better?
I'm also not sure what the "security implications" of storing your data in the registry are. The registry is just as secure as the filesystem - the registry and the filesystem use the same ACL mechanism to protect their data.
If you ARE going to store your user data in a file, you should absolutely put your data in %APPDATA%\CompanyName\ApplicationName at least - that way if two different developers create an application with the same name (how many "Media Manager" applications are there out there?) you won't have collisions.
For me, simple user configuration items and user data is better to be stored in either a simple XML configuration file, a SQLLite db, or a MS SQL Server Compact db. The exact storage medium depends on the specifics of the implementation.
I only use the registry for things that I need to set infrequently and that users don't need to be able to change/see. For example, I have stored encrypted license information in the registry before to avoid accidental user removal of the data.
Using the registry to store data has mainly one problem: It's not very user-friendly. Users have virtually no chance of backing up their settings, copying them to another computer, troubleshooting them (or resetting them) if they get corrupted, or generally just see what their software is doing.
My rule of thumb is to use the registry only to communicate with the OS. Filetype associations, uninstaller entries, processes to run at startup, those things obviously have to be in the registry.
But data that is for use in your application only belongs in a file in your App Data folder. (whiever one of the 3+ App Data folders Microsoft currently wants you to use, anyway)
As each user has directory space in Windows already dedicated to storing application user data, I use it to store the user-level data (preferences, for instance) there.
In C#, I would get it by doing something like this:
Environment.GetFolderPath( Environment.SpecialFolder.ApplicationData);
Typically, I'll store SQLite files there or whatever is appropriate for the application.
If your app is going to be deployed "in the enterprise", keep in mind that administrators can tweak the registry using group policy tools. For example, if firefox used the registry for things like the proxy server, it would make deployment a snap because an admin can use the standard tools in active directory to set it up. If you use anything else, I dont think such things can be done very easily.
So don't dismiss the registry all together. If there is a chance an admin might want to standardize parts of your configuration across a network, put the setting in the registry.
I think Microsoft is encouraging use of isolated storage instead of the Windows registry.
Here's an article that explains how to use it in .Net.
You can find those files in Windows XP under Documents & Settings\\Local Settings\ App Data\Isolated Storage. The data is in .dat files
I would differentiate:
On the one hand there is application specific configuration data that is needed for the app to run, e.g. IP addresses to connect to, which folders to use for what sort of files etc, and non trivial per user settings.
Those I put in a config file, ini format for simple stuff, xml if it gets more complex.
On the other hand there is trivial per user settings (best example: window positions and layout). To avoid cluttering the config files (which some users will want to edit themselves, so few and clearly arranged entries are a must), I like to put those in the registry (with conservative defaults being set in the app if no settings in the registry can be found).
I mainly do it like istmatt sais: I store config files inside the %APPDATA% folder. Usually in %APPDATA%\ApplicationName, I don't like the .NET default of APPDATA%\CompanyName\ApplicationName\Version, that level of detail and complexity is counterproductive for most small to medium sized applications.
I disagree with the example of Marcelo MD of not storing recently used files in the registry. IMO this is exactly the volatile sort of user specific information that can be stored there.
(His example of what not to do is very good, though!)
To me it seems easier to think of what you should NOT put there.
e.g: dynamic data, such as an editor's "last file opened" and per project options. It is really annoying when your app loses sync with the registry (file deletion, system crash, etc) and retrieves information that is not valid anymore, possibly deadlocking the user.
At an earlier job I saw a guy that stored a data transfer completness percentage there, Writing the new values at every 10k or so and having the GUI retrieve this value every second so it could show on the titlebar.

Resources