The peculiarities of testing printer redirection in the terminal sessions
Written by:
Rina Krupskaya,
Senior Tester of QA Direction
Apriorit Inc.
http://www.apriorit.com
Content
1. Printer redirection to the terminal session using Microsoft tools
2. Testing printer redirection to the terminal session
2.1. Functional testing: checking of redirected printers representation, printing in TS
2.2. Non-functional testing of printer redirection to the Terminal Server
Conclusion
1. Printer redirection to the terminal session using Microsoft tools
At the moment when user of the thin client connects to the terminal server using Microsoft Remote Desktop Client the client automatically transmit the information about printers connected to the terminal. Saying terminal server (TS) in this article I mean the server that provides clients with the computing resources - processor time, memory, disk space - for solving their tasks. Terminal server uses this information to automatically create printers in the user session and to establish connection between local and network user printers. If the printer driver name matches the name of the driver installed on the terminal server absolutely or the driver name is present in the file ntprint.inf (there are all printer drivers inbuilt in the installation package of the server software mentioned there), then client printer is created automatically. In this case all documents printed to the local printer connected to the user terminal will be successfully transmitted via the virtual channel over the RDP between the terminal server and thin client.
The automatic creation is impossible when the names on the thin client and the server don’t match. Then the following actions should be performed:
1. Set the possibility of printer utilizing in the Remote Desktop Client (mstsc.exe).
2. If users connect by means of Remote Desktop Web Connection (so called Tsweb that can be downloaded from the Microsoft site) then this possibility should be mentioned in the file default.html (Tsweb package includes it) and the corresponding parameter should be set to “true”:
MsRdpClient.AdvancedSettings2.RedirectPrinters = True
3. In the terminal server settings Administrative Tool (tscc.msc) -> Connections -> RDP-Tcp -> Properties the mapping of the Windows Printer shouldn’t be disabled and the connection of the client printers at the server logon also should be allowed.
4. In the user account settings (Active Directory or Windows Local Security Accounts Manager) -> Environment -> Client Devices, the connection of the client printers also should be enabled.
5. In the Group Policy settings the option “Do not allow client printer redirection” should be enabled.
6. In Services Administrative Tool (services.msc) "Print Spooler Service" should have status “Started”.
7. Printer shouldn’t be "Host-based printer" (so called GDI or Windows-only Printer). "Host-based" printers, such as Hewlett Packard Printers using LIDIL (Lightweight Imaging Device Interface Language), don’t function in the Terminal Services environment (without utilizing the software for printing, developed by different vendors). You can also additionally consult your printer vendor about the support of PCL or PostScript drivers.
8. Printer driver on the terminal server should not use "printer monitor" or so called printer utilities included in the software package of the inexpensive printers. Such utilities (programs) are not intended for the work in the multiuser environment such as Windows Terminal Services.
As you see the procedure described above is not very easy and requires the certain conditions. But people always look for the easier ways to solve their tasks.
Now there exists software that helps to redirect client printers to the session without installation printer drivers on TS. We’ll suppose that we should test such program.
2. Testing printer redirection to the terminal session
Let’s consider the peculiarities of testing client printer redirection to the TS.
To perform all tests described below we suppose that our software is installed on the client and server. Tested software can redirect either all client printers or ones of the specified type (only default, only network, only local). There are local, virtual and network printers on the client machine. We will use Desktop Connection, Remote Application, Remote Desktop Web Connection to connect to the TS.
Let’s formulate the test plan for our application.
2.1. Functional testing: checking of redirected printers representation, printing in TS
1. So, the first testing steps are focused on the printer redirection to the session:
1.1. Default printer redirection;
Expected result: default printer is redirected to the session and represented as the default one. After the session finishes log off is performed.
1.2. Local printer redirection;
Expected result:
- If none of the local printers is default printer then in the session any of the redirected local client printers can be represented as the default one.
- If there is a default printer among the local printers then we check both redirection of the local client printers and the representation of the default printer in the session.
- All local printers are redirected to the session.
Attention: virtual client printers can be also redirected to the session with the local ones.
After the session finishes log off is performed.
1.3. Network printer redirection;
Expected result:
- If none of the network printers is default printer then in the session any of the redirected network client printers can be represented as the default one.
- If there is a default printer among the network printers then we check both redirection of the network client printers and the representation of the default printer in the session.
- All network printers are redirected to the session.
Attention: virtual client printers can be also redirected to the session with the local ones.
After the session finishes log off is performed.
1.4. All printers redirection;
Expected result: check that all printers are redirected to the session and that the default client printer is represented in the session as the default one.
After the session finishes log off is performed.
In the terminal session we will see a picture like this:
1.5. After all printers were redirected to the TS and Printers folder was refreshed in the session (F5), check that all printers are still redirected and the default client printer is still represented as the default one.
2. Attend to the different printers and check their correct redirection to the session.
2.1. Printer with USB interface;
2.2. Printer which is connected to the LPT port;
2.3. Printer which is connected to the COM port;
2.4. Wireless printer.
3. The printer redirection can be changed also “online”, i.e. directly while the session is running and after each session reconnection.
Let’s consider the cases when the session is running.
Number of test | Pre-condition | Steps | Expected result |
3.1. | 1. Session is running 2. Default client printer is redirected to the session. | Without closing the session change the default printer on the client side | New default printer appears in the session, it has replaced the previous default printer |
3.2. | 1. Session is running 2. Default client printer is redirected to the session. | Without closing the session change the redirection of the default printer to the redirection of the network printers on the client side. | 1. If none of the network printers is default printer then in the session any of the redirected network client printers can be represented as the default one. 2. If there is a default printer among the network printers then we check both redirection of the network client printers and the representation of the default printer in the session. 3. All network printers are redirected to the session. |
3.3. | 1. Session is running 2. Default client printer is redirected to the session. | Without closing the session change the redirection of the default printer to the redirection of the local printers on the client side. | 1. If none of the local printers is default printer then in the session any of the redirected local client printers can be represented as the default one. 2. If there is a default printer among the local printers then we check both redirection of the local client printers and the representation of the default printer in the session. 3. All local printers are redirected to the session. |
3.4. | 1. Session is running 2. Network client printers are redirected to the session. | Without closing the session change the redirection of the network printers to the redirection of the local printers on the client side. | 1. If none of the local printers is default printer then in the session any of the redirected local client printers can be represented as the default one. 2. If there is a default printer among the local printers then we check both redirection of the local client printers and the representation of the default printer in the session. 3. All local printers, instead of the network ones, are redirected to the session. |
3.5. | 1. Session is running 2. Network client printers are redirected to the session. | Without closing the session change the redirection of the network printers to the redirection of the default printer on the client side. | 1. If there was a default printer among of the network printers then this default network printer will be redirected to the session and represented as the default one. 2. If there was not a default printer among the network printers then the client default printer, instead of the network printers, is redirected to the session and represented as the default one. |
3.6. | 1. Session is running 2. Local client printers are redirected to the session. | Without closing the session change the redirection of the local printers to the redirection of the default printer on the client side. | 1. If there was a default printer among of the local printers then this default local printer will be redirected to the session and represented as the default one. 2. If there was not a default printer among the local printers then the client default printer, instead of the local printers, is redirected to the session and represented as the default one. |
3.7. | 1. Session is running 2. Local client printers are redirected to the session. | Without closing the session change the redirection of the local printers to the redirection of the network printers on the client side. | 1. If none of the network printers is default printer then in the session any of the redirected network client printers can be represented as the default one. 2. If there is a default printer among the network printers then we check both redirection of the network client printers and the representation of the default printer in the session. 3. All network printers, instead of the local ones, are redirected to the session. |
The described above tests can be used and for the checking redirection of the client printers after the each session reconnection (log offdisconnect).
Log off — shutting down all applications working in the terminal session, releasing all system resources, allocated for the current session (i.e. session is completely removed from the server);
Disconnect — the window of terminal session is shut down on the client machine, the session is completely stored on the server with all applications running. After the repeat connection of this user to the server the work in the session is fully renewed (Attention! If you finish the work of the remote desktop client by clicking the red cross at the upper right corner of the client application then just Disconnect will be performed).
4. The special attention should be paid on the printer state. We simulate different printer states on the client machine and check how the state is shown in the session.
4.1. Ready printer;
4.2. Paused printer;
4.3. Offline printer;
5. The next test is aimed at the starting multiple sessions by one user (here we assume that the TS itself supports such option).
Let’s consider the case when a user starts two terminal sessions with redirection of all printers to the session.
Expected result: Two sessions contain the similar information. All client printers are redirected to the sessions. We can see duplicated client printers in each session, the only difference is in the unique session number (in the example on the Picture they are 23 and 24).
6. This group of tests is to check the correct redirection of the client printers to TS after the TS restart or client machine restart. You may think: what so unusual can happen after the usual restart? In truth I came across the situations when client printers were not redirected to the TS or were redirected but not in a body.
6.1. Redirection of all client printers after server restart.
Expected result: all printers are redirected to the session and default client printer is shown as the default one.
6.2. Redirection of all client printers after client restart.
Expected result: all printers are redirected to the session and default client printer is shown as the default one.
7. This test is to check the redirection of all printers from different users - administrator, user, guest – to the TS.
Expected result: in all cases irrespective of what user is logged in we assume that all printers are redirected to the session and default client printer is shown in the session as the default one.
8. After we have checked that the client printers are correctly redirected to the TS, we can proceed to the main functionality of our application – to printing.
Here are some points of our keen interest.
8.1. Test of picture printing
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- picture is printed to the client printer and corresponds to the original picture.
8.2. Test of the printing of the text with different fonts.
This test is very important as far as there can be special fonts installed on the server and not installed on the client. As the result we just won’t see the corresponding part of the text or see it but not in the readable form.
For example such specific fonts are APAPA.TTF, ACUTATR_.TTF, AnnabelScript.ttf, apicture.ttf, ARIALUNI.TTF.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- all fonts are transmitted together with the printed document to the client printer and correspond the original fonts.
8.3. Test of the formula, diagram, table printing.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- formula/diagram/table is printed on the client printer and corresponds the original one.
8.4. Test of bar code printing.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- bar code is printed on the client printer and corresponds the original one.
8.5. Test of printing from various account applications.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- document sent from account program(s) is printed on the client printer and corresponds the original one.
8.6. Test of watermark printing.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- watermark is printed on the client printer and corresponds the original one.
8.7. Test of printing of documents of different types: *.doc, *.xls, *.ppt, *.txt, *.pdf, *.html, *.mmap (MindManager), *.jpg etc.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- the content of all documents of all types is printed on the client printer and corresponds the original one.
I tried to cover the main test scenarios of printing, in truth there are a huge number of them.
9. On the next stage we check how main printing options work.
9.1. Printing of all pages of the document;
9.2. Printing of selected pages;
9.3. Printing of several copies of the document;
9.4. Printing of the current page;
9.5. Two-sided printing;
9.6. Printing to the file;
9.7. Printing of the even-odd pages;
9.8. Printing of the selected range;
9.10. Portrait oriented printing;
9.11. LandScape oriented printing;
Etc.
It’s clear that the more properties we cover the higher quality of final product we get.
10. The next test is to check the printing tasks form the queue. To perform this test we send several documents to the one printer.
Expected result:
- the task has not hung in the pool;
- printer icon is shown not in the session but on the client machine;
- tasks with queue status turn to the In Progress status one by one with following printing to the client printer.
11. As you know each user has his access rights for TS. Let’s consider two categories of users:
User1 is a member of the groups:
- administrators;
- remote desktop users.
User2 is a member of the groups:
- users;
- remote desktop users.
We assume that each user has some printers. When both users simultaneously redirect their printers to the one TS and open Printers folder in their sessions they see different data.
Expected result:
User1 sees his printers, server printers (if there are any) and printers of User2.
User2 sees only his printers and server printers (if there are any).
Remember: User1 can’t print to the User2 printers.
2.2 Non-functional testing of printer redirection to the Terminal Server
12. Ideally all mentioned test scenarios should be performed on the different configurations. Configuration list depends on the requirements, of cause.
The most widely spread configuration looks as follows:
Client | Server | ||||||||||
2003 sp2 x32 | 2003 sp2 x64 | 2008 sp2 x32 | 2008 sp2 x64 | Vista sp2 x32 | Vista sp2 x64 | Xp sp3 x32 | Xp sp3 x64 | Win7 x32 | Win7 x64 | Win Server 2000 | |
XP sp3 x32, RDP 6.0 |
|
|
|
|
|
|
|
|
|
|
|
XP sp3 x32, RDP 7.0 |
|
|
|
|
|
|
|
|
|
|
|
XP sp2 x64 |
|
|
|
|
|
|
|
|
|
|
|
Vista sp2 x32 |
|
|
|
|
|
|
|
|
|
|
|
Vista sp2 x64 |
|
|
|
|
|
|
|
|
|
|
|
Win7 x32 |
|
|
|
|
|
|
|
|
|
|
|
Win7 x64 |
|
|
|
|
|
|
|
|
|
|
|
As you understand it will take a huge time to fill in all cells of this table. In practice we just define priorities and adhere to them.
Priorities depends on the requirements. If all above mentioned OS are present in the specification then we base on the most obvious configurations. For example the configuration TS 2003€8 with clients XP sp3 x32 RDP 6.0, Vista sp2 x32, Win7 x32 is the highest priority and it should be passed in the first turn for the first release version. Then we can gradually cover other platforms.
I want to admit one more time that such priorities should be discussed and agreed with the customer.
13. Let’s consider the situation when our software has some limitations on the number of users to redirect printers to the terminal session. This limitation can depends on the license key.
For example only 5 users, who work simultaneously on the Terminal Server, can redirect their printers to the session.
Let’s take a look at the possible test scenarios:
Number of test | Pre-condition | Steps | Expected result |
13.1. | Five users with our software installed have connected to the one TS. | The sixth user with our software installed connects to the same TS. | 1. Printers of the first five users are redirected to the session. 2. Printers of the sixth user are not redirected to the session. |
13.2. | Five users with our software installed have connected to the one TS. | The sixth user with our software installed connects to the same TS. | 1. Printers of the first five users are redirected to the session. 2. Printers of the sixth user are not redirected to the session. |
|
| One of the first five users logs off, for example the first one. | 1. Printers of the first user are not shown on the TS after he logged off. 2. Printers of the 2nd, 3rd, 4th, 5th users are shown on the TS as before. 3. Printers of the 6th user are not redirected to the session. |
|
| The 6th user logs off. | The session of the 6th user is finished. |
|
| The 6th user connects to the TS. | Printers of the 2nd, 3rd, 4th, 5th, 6th users are shown on the TS. |
|
| The 1st user connects to the TS. | 1. Printers of the 2nd, 3rd, 4th, 5th, 6th users are shown on the TS. 2. Printers of the 1st user are not redirected to the session. |
13.3. | Three users with our software installed connect to the one TS. Let them be user1, user2, user3. | user4 and user5 connect to the same TS, they don’t have our software installed. | 1. Printers of user1, user2, user3 are redirected to the TS. 2. Printers of user4, user5 are not redirected. |
|
| user6 with our software installed connects to the same TS. | Printers of user1, user2, user3, user6 are redirected to the TS. |
13.4. | Five users, who don’t have our software installed, connect to the one TS. Let them be user1, user2, user3, user4, user5. | Five more users connect to the same TS, they have our software installed. Let them be user6, user7, user8, user9, user10. | Printers of user6, user7, user8, user9, user10 are redirected to the TS. |
14. We should also remember about the performance testing.
So we should measure the speed of printing in the terminal session when our software is installed. But how can we define the optimal performance level for our application? Sometimes the numbers are specified by the customer or project manager but much more often we have to define performance bounds ourselves.
Using the common sense we can say that the application has bad performance if it takes too long to print the small document.
We can also measure the printing speed using Microsoft RDPRemote AppRemote Desktop Web Connection and checking the Redirect printers option – here we assume that the drivers of client printer are installed on the TS and our software is not installed. This neutral situation gives us some number to compare with the printing speed of the same documents with our software. It’s obvious that the speed with our application should not be less than the corresponding speed without our software, and ideally it should be greater.
15. Load testing.
15.1. Let’s consider how the main functionality will work in the high load conditions. For example, continuous printing during several hours/days. To implement this we can develop the small script to send document from the session to the client printer in the loop.
Expected result: the test is for check the performance (after the n-th hour/day the performance should not become worse). And also, the printing is possible and the corresponding task is not hung in the pool.
15.2. Different users connect to the same TS from the one client machine and send a document to the one client printer. Here we can check how the printing from the multiple users to the one printer is performed.
Expected result:
- printing is possible for two users;
- the printer icon appears not in the session but on the client PC;
- the tasks are not hung in the pool;
- the tasks that have queue status pass into In Progress status one by one with the subsequent printing to the client printer.
16. Stress testing.
16.1. Perform session disconnect while the printing is going.
Expected result: printing is stopped; after the next connection to the server all client printers are redirected correctly and printing is possible. Client printers are not duplicated in the session.
16.2. Perform session log off while the printing is going.
Expected result: printing is stopped; after the next connection to the server all client printers are redirected correctly and printing is possible.
16.3. Restart or reset the TS while the client printers are redirected to the session. As practice shows after that manipulation either the client printers are duplicated in the session or the default printer is shown incorrectly. So we check if all printers are not duplicated and the default printer is shown correctly.
16.4. During the redirection and/or printing process the network cable can be plugged off or the network connection can be switched off.
Expected result: printing is stopped; after the next connection to the server all client printers are redirected correctly and printing is possible.
Conclusion
In this article I mentioned the most important moments of testing of printer redirection to the terminal session.
Of cause you can say that not everything is taken into account and you can propose much more testing scenarios. It’s good and it’s true – I based only on my experience and peculiarities of my projects.
I wish you a lot of successful releases! And let your products will be of high quality!
