The max-filtered bandwidth and min-filtered RTT estimates form BBR's model of the network pipe. Conversely it uses a windowed min-filter to estimate the round trip propagation delay. On the arrival of each ACK, BBR derives the current delivery rate of the last round trip, and feeds it through a windowed max-filter to estimate the bottleneck bandwidth. On today's high-speed long-haul links using commodity switches with shallow buffers, loss-based congestion control has abysmal throughput because it over-reacts to losses caused by transient traffic bursts.īBR creates an explicit model of the network pipe by sequentially probing the bottleneck bandwidth and RTT. On today's Internet, loss-based congestion control causes the infamous bufferbloat problem, often causing seconds of needless queuing delay, since it fills the bloated buffers in many last-mile links. While this worked well for many years, loss-based congestion control is considered -by BBR authors- to be out-dated in today's networks. The Internet has predominantly used loss-based congestion control, relying on packet loss as the signal to slow down. This release adds another TCP congestion control algorithm: BBR (Bottleneck Bandwidth and RTT). For more details, see Brendan Gregg's blog postĬode: commit, commit, commit, commit, commit, commitġ.4. This release includes the necessary infrastructure to allow BPF programs to be atached to hardware and software perf events, which in turn allows the kernel to profile via timed sampling and summarize stack traces. Recommended LWN articles: Virtually mapped kernel stacks, Virtually mapped stacks 2: thread_info strikes back This release allows to map the kernel stacks in virtual memory, which makes easier to allocate stacks under memory pressure, provides protection against stack overflows, and other benefits. The Linux kernel has always mapped the memory used by kernel stacks directly in the kernel memory, an approach that makes harder to allocate stacks under memory load and proves no protection against stack overflows. Virtually mapped stacks for more secure and reliable stack handling It also adds copy-on-write support for data: instead of overwriting data, it copies data to a new location.Īll these features are a huge chunk of new experimental functionality with new on-disk format features and internal infrastructure.ġ.2. That is, support for cp -reflink=always, the ability to deduplicate data, and the ability to unshared data through the FALLOC_FL_UNSHARE fallocate(2) interface. This release adds the ability to share data extents between different files. This release adds several key features to the XFS file system, based on the reverse mapping work introduced in the past release. ![]() Shared data extents + copy-on-write support on XFS Voltage, current regulators, power capping, power supplyġ.1.ACPI, EFI, cpufreq, thermal, Power Management.Tablets, touch screens, keyboards, mouses. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |