/strata

Strata: A Cross Media File System

Primary LanguageC

Strata: A Cross Media File System

This repository is no longer maintained; Strata is now under the umbrella of the Assise project.

Strata is a research prototype file system, presented in SOSP 2017 (Strata).

Strata is developed and tested on Ubuntu 16.04 LTS, Linux kernel 4.8.12 and gcc version 5.4.0.

This repository contains initial source code and tests. Benchmarks will be released soon. As a research prototype, Strata has several limitations, described in the limitations section.

To run NVM emulation, your machine should have enough DRAM for testing. Kernel will reserve the DRAM for NVM emulation. Strata requires at least two partitions of NVM: operation log (1 - 2 GB) and NVM shared area (It depends on your test. I recommend to use more than 8 GB at least).

Building Strata

Assume current directory is a project root directory.

1. Change memory configuration
./utils/change_dev_size.py [dax0.0] [SSD] [HDD] [dax1.0]

This script does the following:

  1. Opens libfs/src/storage/storage.h
  2. Modifiesdev_size array values with each storage size (the same as in your grub conf, see the running Strata section) in bytes.
    • dev_size[0]: could be always 0 (not used)
    • dev_size[1]: dax0.0 size
    • dev_size[2]: SSD size : just put 0 for now
    • dev_size[3]: HDD size : put 0 for now
    • dev_size[4]: dax1.0 size
2. Build kernel
cd kernel/kbuild
make -f Makefile.setup .config
make -f Makefile.setup
make -j
sudo make modules_install ; sudo make install

This step requires reboot your machine after installing the new kernel.

3. Build dependent libraries (SPDK, NVML, JEMALLOC)
cd libfs/lib
git clone https://github.com/pmem/nvml
git clone https://github.com/pmem/syscall_intercept.git
make

tar xvjf jemalloc-4.5.0.tar.bz2
cd jemalloc-4.5.0
./autogen
./configure
make

For SPDK build errors, please check a SPDK website (http://www.spdk.io/doc/getting_started.html)

For NVML build errors, please check a NVML repository (https://github.com/pmem/nvml/)

4. Build Libfs
cd libfs
make
5. Build KernelFS
cd kernfs
make
cd tests
make

Running Strata

1. Setup NVM (DEV-DAX) emulation

Strata emulates NVM using a physically contiguous memory region, and relies on the kernel NVDIMM support.

You need to make sure that your kernel is built with NVDIMM support enabled (CONFIG_BLK_DEV_PMEM), and then you can reserve the memory space by booting the kernel with memmap command line option.

For instance, adding memmap=16G!8G to the kernel boot parameters will reserve 16GB memory starting from 8GB address, and the kernel will create a pmem0 block device under the /dev directory. Adding GRUB_CMDLINE_LINUX="memmap=16G!4G, 4G!20G" will add a pmem0 and pmem1.

Details are available at: http://pmem.io/2016/02/22/pm-emulation.html

This step requires rebooting your machine.

2. Use DEV-DAX emulation
cd utils
sudo ./use_dax.sh bind

This instruction will change pmem emulation to use dev-dax mode.

e.g., /dev/pmem0 -> /dev/dax0

To rollback to previous setting,

sudo ./use_dax.sh unbind
3. Setup storage size

This step requires rebuilding of Libfs and KernFS.

TODO: Some instructions to setup storage size (by a script or manually)
4. Setup UIO for SPDK
cd utils
sudo ./uio_setup.sh linux config

To rollback to previous setting,

sudo ./uio_setup.sh linux reset
5. Formatting storages
cd libfs
sudo ./bin/mkfs.mlfs <dev id>

dev id is a device identifier used in Strata (hardcoded).
1 : NVM shared area
2 : SSD shared area
3 : HDD shared area
4 : Operation log of processes

If you encounter an error message, "mmap invalid argument", it means kernel does not allow mmap for NVM emulation. Usually, incorrect (or unaligned) setting of storage sizes (at step 3) causes the problem. Please make sure that your storage size is correct in "libfs/src/storage/storage.h"

6. Run KernelFS
cd kernfs/tests
make
sudo ./run.sh kernfs
7. Run testing problem
cd libfs/tests
make
sudo ./run.sh iotest sw 2G 4K 1 #sequential write, 2GB file with 4K IO and 1 thread

Strata configuration

1. LibFS configuration

In libfs/Makefile, search MLFS_FLAGS as keyword

MLFS_FLAGS = -DLIBFS -DMLFS_INFO
#MLFS_FLAGS += -DCONCURRENT
MLFS_FLAGS += -DINVALIDATION
#MLFS_FLAGS += -DKLIB_HASH
MLFS_FLAGS += -DUSE_SSD
#MLFS_FLAGS += -DUSE_HDD
#MLFS_FLAGS += -DMLFS_LOG

DCONCURRENT - allow parallelism in libfs
DKLIB_HASH - use klib hashing for log hash table
DUSE_SSD, DUSE_HDD - make LibFS to use SSD and HDD

2. KernelFS configuration
#MLFS_FLAGS = -DKERNFS
MLFS_FLAGS += -DBALLOC
#MLFS_FLAGS += -DDIGEST_OPT
#MLFS_FLAGS += -DIOMERGE
#MLFS_FLAGS += -DCONCURRENT
#MLFS_FLAGS += -DFCONCURRENT
#MLFS_FLAGS += -DUSE_SSD
#MLFS_FLAGS += -DUSE_HDD
#MLFS_FLAGS += -DMIGRATION
#MLFS_FLAGS += -DEXPERIMENTAL

DBALLOC - use new block allocator (use it always)
DIGEST_OPT - use log coalescing
DIOMERGE - use io merging
DCONCURRENT - allow concurrent digest
DMIGRATION - allow data migration. It requires turning on DUSE_SSD

For debugging, DIGEST_OPT, DIOMERGE, DCONCURRENT is disabled for now

Limitations

  1. KernelFS is currently implmented in user-level.
  2. Leases are not fully implemented.
  3. A directory could contain up to 1000 files.
  4. mmap is not supported yet.
  5. Benchmarks are not fully tested in all configurations. Working configurations are described in our paper.
  6. There are known bugs in fork.