Distributed Systems

Distributed Systems Lab 4: MKDIR, UNLINK, and Locking

Due: Friday, Dec 2nd, 11:59pm

Introduction

In this lab, you will continue to add functionality to your file server. You will:

• Add handlers for the MKDIR and UNLINK FUSE operations,
• Add simple locking. Locking is required to ensure that concurrent modifications to the same file or directory occur one at a time.

Getting Started

First, merge your solution to Lab 3 with the new code for Lab 4. There are no substantive differences in our code between the labs, except for testers. We have also provided stubs for the FUSE functions you need to implement in order to make your job easier. You can use the following commands to get the Lab 4 sources; see the directions for Lab 2 for more background on git.

% cd lab

% git commit -am 'my solution to lab3'

Created commit ...

% git pull

remote: Generating pack...

...

% git checkout -b lab4 origin/lab4

Branch lab4 set up to track remote branch refs/remotes/origin/lab4.

Switched to a new branch "lab4"

% git merge lab3

As before, if git reports any conflicts, edit the files to merge them manually, then run git commit -a. Since you are building on the previous labs, ensure the code in your Lab 3 directory passes all tests for Labs 1, 2, and 3 before starting in on this lab.

The rest of this lab has two parts.

Part 1: MKDIR,UNLINK

Part 1: Your Job

Your job in part 1 is to handle the MKDIR and UNLINK FUSE operations. This should be a straightforward extension of your Lab 3 code. Make sure that when you choose the inumber for a new directory created with MKDIR, that inumber must have its most significant bit set to 0 (as explained in Lab 2, unless you changed the way YFS tells files and directories apart). When you're done with Part 1, the following should work:

% ./start.sh

% mkdir yfs1/newdir

% echo hi > yfs1/newdir/newfile

% ls yfs1/newdir/newfile

yfs1/newdir/newfile

% rm yfs1/newdir/newfile

% ls yfs1/newdir

% ./stop.sh

If your implementation passes the test-lab-4-a.pl script, you are done with part 1. The test script creates a directory, creates and deletes lots of files in the directory, and checks file and directory mtimes and ctimes. Note that this is the first test that explicitly checks the correctness of these time attributes. A create should change both the parent directory's mtime and ctime. Here is a successful run of the tester:

% ./start.sh

% ./test-lab-4-a.pl ./yfs1

mkdir ./yfs1/d3319

create x-0

delete x-0

create x-1

checkmtime x-1

...

delete x-33

dircheck

Passed all tests!

% ./stop.sh

Part 2: Locking

Next, you are going to ensure the consistency of your file system when many clients simultaneously perform file system operations on the same file system image via different yfs_client processes. Your current implementation does not handle concurrent operations correctly. For example, your yfs_client's create method probably reads the directory's contents from the extent server, makes some changes or additions, and stores the new contents at the extent server. Suppose two clients issue simultaneous CREATEs for different file names in the same directory via different yfs_client processes. Both yfs_client processes might fetch the old dir contents at the same time and each inserts the newly created file for its client and writes back the new dir contents. As a result, only one of the file would be present in the dir in the end. The correct answer, however, is for both files to exist. The CREATE example is just one of the "race conditions". Many others exist: e.g. concurrent CREATE and UNLINK, concurrent MKDIR and LOOKUP, etc.

To fix the race conditions, the yfs_client must use locks to ensure that the two operations that access the same file or directory happen one at a time. For example, a yfs_client would acquire a lock before starting the CREATE, and only release the lock after finishing the write of the new information back to the extent server. If there are concurrent operations, the locks force one of the two operations to delay until the other one has completed. Because each yfs_client can run as a separate process on a different machine, all yfs_clients have to acquire locks from the same lock server. Now you can see why the lock server implementation from Lab 1 comes in handy!

Part 2: Your Job

Your job is to implement locking for yfs_client to ensure that concurrent operations from different yfs_clients proceed correctly. The testers for this part of the lab are test-lab-4-b and test-lab-4-c (The source files are test-lab-4-b.c and test-lab-4-c.c). The testers take two directories as arguments and issue concurrent operations in the two directories and check that the results are consistent with the operations executing in some serial order. Here's a successful execution of the testers:

% ./start.sh

% ./test-lab-4-b ./yfs1 ./yfs2

Create then read: OK

Unlink: OK

Append: OK

Readdir: OK

Many sequential creates: OK

Write 20000 bytes: OK

Concurrent creates: OK

Concurrent creates of the same file: OK

Concurrent create/delete: OK

Concurrent creates, same file, same server: OK

test-lab-4-b: Passed all tests.

% ./stop.sh

%

% ./start.sh

% ./test-lab-4-c ./yfs1 ./yfs2

Create/delete in separate directories: tests completed OK

% ./stop.sh

If you try this before you add locking, it will fail at "Concurrent creates" test in test-lab-4-b.

After you are done with part 2, you should also test with test-lab-4-a.pl to make sure you didn't break anything. You might also test with test-lab-4-b with the same directory for both arguments, to make sure you handle concurrent operations correctly with only one server before you go on to test concurrent operations in two servers.

Part 2: Detailed Guidance

·        What to lock?

You must choose what the locks refer to. At one extreme you could have a single lock for the whole file system, so that operations never proceed in parallel. At the other extreme you could lock each entry in a directory, or each field in the attributes structure. Neither of these is a good idea! A single global lock prevents concurrency that would have been okay, for example CREATEs in different directories. Fine-grained locks have high overhead and make deadlock likely, since you often need to hold more than one fine-grained lock.

Your best bet is to associate one lock with each file handle. Use the file or directory's inumber as the name of the lock (i.e. pass the inumber to acquire and release). The convention should be that any yfs_client operation should acquire the lock on the file or directory it uses, perform the operation, finish updating the extent server (if the operation has side-effects), and then release the lock on the inumber. You must be careful about releasing the locks in all circumstances upon return from yfs_client operation.

You'll use your lock server from Lab 1. Our original template for the yfs_client constructor that we gave you in Lab 2 included the destination address of a lock server, so it should be very easy to add a lock_client object to the yfs_client and simply call its acquire and release methods.

·        Things to watch out for:

This is the first lab that creates files using two different YFS-mounted directories. If you were not careful in earlier labs, you may find that the components that assign inumbers for newly-created files and directories choose the same identifiers. One possible way to fix this may be to seed the random number generator differently depending on the process's pid.

This is also the first lab that writes arbitrary data to the file, rather than null-terminated C-style strings. If you used the standard std::string constructor in fuse.cc to create a string to pass to your yfs_client, (i.e., std::string(buf)), you will get odd errors when there are characters equal to '\0' in the buffer. Instead, you should use a different constructor that allows for char buffers of arbitrary data: std::string(buf, size).

Handin procedure

You will need to email your completed code (without binaries) as a gzipped tar file to ds-assignment@mpi-sws.org by the deadline stated at the top of the page. To do this, switch to the source directory and execute these commands:

% tar czvf MATR1-MATR2-lab4.tgz lab/

That should produce a file called [MATR1-MATR2]-lab4.tgz in that directory, where MATR1 and MATR2 are the matriculation numbers of the team members. Attach that file to an email and send it to the address above with the subject "Assignment 4 - LastName1 LastName2".

You will receive full credit if your software passes the same tests we gave you when we run your software on our machines.

Questions or comments regarding this course? Please use the general course mailing list or the teaching staff mailing list.

Top // Distributed Systems //