Scalable Hazard Pointer Synchronous Reclamation Beyond Concurrency TS2 – Maged Michael Cohort Structures Revisited Hazard Pointer Synchronous Reclamation Beyond Concurrency TS2 – Maged Michael class microbenchmarking. • 2020: One user: High frequency of retiring objects. Unsharded list grew out of control. • 2020: Sharded cohorts without reclamation under lock. Fast. Scalable. Robust. Quasi-Synchronous

“All software construction involves essential tasks, the fashioning of the complex conceptual structures that compose the abstract software entity, and accidental tasks, the representation of the abstract the essential parts of the software task, those concerned with fashioning abstract conceptual structures of great complexity. I suggest: • Using rapid prototyping as part of a planned iteration in establishing which address the essence of the software problem, the formulation of these complex conceptual structures”Attack the essence of the software problem • Buy versus build • Requirements refinement and

written in other languages, typically C or C++. FFI allows Rust to call functions and use data structures defined in other languages, and vice versa. 利用 Rust 重塑移动应用开发 Photo / image / chart Rust FFI

hood • Representing the store • Solving constraints • Guiding the analysis • Immutable data structures? • Modeling loops?From source to values Stack struct S { int *p; }; int f() { S a[10];

shared_ptr s1: control_block: atomic ref_count = 1; … control_block* ctrl; T12 Daniel Anderson -- danielanderson.net std::shared_ptr (the basics) T* ptr; shared_ptr s1: control_block: control_block* shared_ptr s2: control_block* ctrl; atomic ref_count = 2; …13 Daniel Anderson -- danielanderson.net std::shared_ptr (the basics) control_block: T T* ptr; control_block* ctrl; atomic … shared_ptr s2:14 Daniel Anderson -- danielanderson.net std::shared_ptr (the basics) control_block: T atomic ref_count = 0; …15 Daniel Anderson -- danielanderson.net std::atomic16

std::shared_ptr(block) Control Block T T T T T …my::shared_ptr Control Block T T T T T T … int ref_cnt; int size; my::shared_ptr(block)I will write it myself ™ Control Block T T T T T T … templatecontrol block // Cast and initialize each T // return my::shared_ptr(block); } // FixMe: What

done_as_error(sender, error) → sender … commutes the done signal into an error.19 SENDER/RECEIVER CONTROL FLOW20 BASIC LIFETIME OF AN ASYNC OPERATION SCHEDULER schedule SENDER Implementation details STRUCTURED PROGRAMMING Structured control flow constructs have a single entry and a single exit, permitting them to be treated like a black box. goto is an unstructured control flow construct conditional checks if there is a pending completion. If so, it completes it. The constructor registers the control handler, and the destructor unregisters it.104 [[nodiscard]] auto event() const; }; auto c

GIT is a version control system that is used for software development and other version control tasks. As a distributed revision control system it is aimed at speed, data integrity, and support for

chain • Pros • Iptables is widely adopted in popular Linux distributions • Cons • O(N^2) in control plane / O(N) in data plane • Poor in scheduling algorithm • Iptables rules are difficult to debug organized in hash table • IPVS DNAT • conntrack/iptables SNAT • Pros • O(1) time complexity in control/data plane • Stably runs for two decades • Support rich scheduling algorithm • Cons • Performance

traffic ● Authorize in mesh traffic ● Summary #IstioCon Istio Architecture Connect, secure, control, and observe services. #IstioCon Security Architecture #IstioCon Bookinfo architecture without will help you on board to Istio ● Strict mTLS if possible ● Secure your ingress ● Enable Access Control to your services via Istio authorization policy Istio will Secure your microservices for you!