Cs704 d distributedschedulingetc.
Download as PPTX, PDF1 like489 views
The document discusses different approaches and algorithms for scheduling tasks in distributed systems, including task assignment, load balancing, and load sharing. It covers topics like static vs dynamic scheduling, centralized vs distributed approaches, and policies for load estimation, process transfer, location selection, state information exchange, and priority assignment. Key aspects of designing scheduling algorithms like stability, scalability, and fault tolerance are also mentioned.
Cs704 d distributedschedulingetc.
- 1. Debasis Das
- 2. Scheduling CS 704D Advanced OS 2
- 3. Scheduling Approaches Task assignment approach Schedule the tasks in a user submitted process to suitable nodes for improving performance Load balancing approach Tasks are distributed to nodes to balance workload of the nodes Load sharing approach Ensures no node is idle when a process may wait for a processor CS 704D Advanced OS 3
- 4. Desirable Features of Good Scheduling Algorithm No A priori knowledge about the processes Dynamic in nature Quick decision making capability Balanced system performance and scheduling overhead Stability Scalability Fault tolerance Fairness of service CS 704D Advanced OS 4
- 5. Task Assignment Approach The Basic Idea Assumptions (so that tasks can be assigned optimally) The process is already split into tasks, that are broken at natural boundaries and data transfers can be minimized between tasks Computation required by a task and speed of each processor are known Cost of execution of the task at each node is known IPC costs between pairs of processes are known Other constraints, such as resource requirements, resources at each node, precedence among tasks etc are known Reassignment of tasks is , generally, not possible CS 704D Advanced OS 5
- 6. Task Assignment Approach The optimization to be done Minimize IPC costs Quick turn around for the process High degree of parallelism Efficient utilization of system resources, in general CS 704D Advanced OS 6
- 7. Load Balancing Approach Static vs. dynamic Deterministic vs. probabilistic Centralized vs. Distributed Cooperative vs. Non-cooperative CS 704D Advanced OS 7
- 8. Load Balancing Taxonomy CS 704D Advanced OS 8 Load balancing Algorithms Deterministic Static Dynamic Centralized Distributed Non-cooperativeCooperative Probabilistic
- 9. Static vs. Dynamic Static: Uses average behavior of system, ignores current state Simple, no need to process system state information But does not adjust to current situation Dynamic: Reacts to system changes dynamically Responds to system state, avoids unnecessary poor performance Provides better performance , but complex as system state information needs to be collected and processed CS 704D Advanced OS 9
- 10. Deterministic Vs. Probabilistic Deterministic processes allocate tasks based on known properties of nodes and the processes to be scheduled A probabilistic algorithm uses information regarding static attributes of the system such as number of nodes, processing capability at each node, network topology and formulate simple placement rules Deterministic algorithms are difficult to optimize and costs more to implement CS 704D Advanced OS 10
- 11. Centralized vs. Distributed A centralized server node assigns tasks based on state information collected at that node. All other processor nodes send information to that node and keep it updated. Quite efficient But, single point of failure is the central node Solution is to replicate k+1 servers if k faults are to be serviced. Issue of consistency arises If a delay of a few seconds are acceptable re- instantiation can be used CS 704D Advanced OS 11
- 12. Distributed Scheme K physically distributed entities Each is considered a local controller, runs concurrently and asynchronously with other entities Entities make decisions based on system wide objective function In a fully distributed system all N nodes are the scheduling entities scheduling local as well as acceptance of remote processes CS 704D Advanced OS 12
- 13. Cooperative vs. Non-cooperative In non-cooperative mode entities make decisions independent of other nodes In cooperative mode the entities cooperate to arrive at a decision Cooperative algorithms are thus complex and has higher overheads Cooperative algorithms have better stability though CS 704D Advanced OS 13
- 14. Design issues Load estimation policy Process transfer policy State information exchange policy Location policy Priority assignment policy Migration limiting policy CS 704D Advanced OS 14
- 15. Load estimation Policy Parameters that are time dependent and node dependent are Number of processes on the node at the time of estimation Resource demands of these processes Instruction mixes of these processes Architecture and speed of the node’s processor CS 704D Advanced OS 15
- 16. Load Estimation Methods A simple measure could be number of processes executing at the node That may not be accurate as the load really depends on the remaining service time of these processes, some methods are Memoryless method: assumes remaining time is same for all, irrespective of time used by these processes, load estimate is thus just the number of processes method Pastrepeats: remaining service time is equal to the time used by it so far Distribution method: if the distribution of services times is known then the remaining service time is expected remaining service time conditioned by time already used CS 704D Advanced OS 16
- 17. Process Transfer Policy Threshold decides if a process will be transferred Static policy: predefined threshold, no exchange of state information is required to decide the threshold Dynamic policy: the threshold is the product of average workload of all nodes and a predefined constant Ci. Ci is proportional to the processing capability of ni relative to the processing capability of other nodes. State information needs to be exchanged to keep determining the load CS 704D Advanced OS 17
- 18. Process Transfer-Threshold Single threshold policy: rigid single threshold can lead to instability as if the local load is just below threshold and a process is accepted, the accepted process increases the load beyond the threshold Double threshold policy: The threshold is a band creating a high watermark and a low watermark to decide transfers (overloaded, high mark, normal load, low mar, under-loaded) Necessary that the newer processes coming messages should not affect the local load significantly CS 704D Advanced OS 18
- 19. Thresholding CS 704D Advanced OS 19 Overloaded Underloaded Overloaded Underloaded Normal Threshold High mark Low mark
- 20. State Information Exchange Policy Periodic Broadcast: lot of communication load, does not scale well Broadcast when state changes: further refinement is to transmit state information only when state changes to under loaded or overloaded On-demand exchange: request state when the load changes to under loaded/ overloaded from normal ( note: only nodes that are not normal need to convey information) Exchange by polling: When needed a node can poll others CS 704D Advanced OS 20
- 21. Location Policy Threshold: select nodes at random, check if placing the process will overload it Shortest: a pre-decided number of nodes are polled, process transferred to one with the least load Bidding: request bids for a process to be migrated, nodes bid for it. Best bidder, in terms of fastest time, cheapest processing, best price/performance Pairing: low load and high load nodes are paired, there may be requests seeking pairing CS 704D Advanced OS 21
- 22. Priority Assignment Policy Selfish: Local processes get higher priority Altruistic: remote processes get higher priority Intermediate: Local processes get higher priority if the number is more than or equal to remote processes number. Otherwise remote processes are prioritized CS 704D Advanced OS 22
- 23. Priority Assignment Comparisons Selfish: worst response time for remote processes, higher penalty I remote process arrives at highly loaded node, lower otherwise Altruistic: best response time, local processes may face penalty Intermediate: response time performance is close to that of altruistic policy yet the local processes do not suffer as much CS 704D Advanced OS 23
- 24. Migration Limiting Policy Uncontrolled: Local and remote processes treated the same way. migration may happen any number of times, can cause instability Controlled: A migration count is used. Many designers feel migration is expensive and thus it is limited to 1. Some prefer limiting it to a small value k, determined statically or dynamically CS 704D Advanced OS 24
- 25. Load Sharing Approach Load sharing may be better than load balancing, load fluctuates and the overhead of exchanging state information is quite a bit. It might be better to make sure no nodes idle when any node has more than one process Priority assignment and migration limiting policies are same as with load balancing approach CS 704D Advanced OS 25
- 26. Load Sharing Design issues Load estimation Policies Process transfer policies Location policies Sender initiated location policies Receiver ini8tiated location policies State Information Exchange policies Broadcast when state changes Poll when state changes CS 704D Advanced OS 26
- 27. Load Estimation Policies Simplest load estimation required to find only if a node is idle or overloaded by counting number of processes at a node Most modern systems have some processes running all the time at any node. A better estimator may be the CPU utilization CS 704D Advanced OS 27
- 28. Process Transfer Policies All or nothing strategy Fix threshold at 1 A node will accept a process if it has no processes, an will try to send away a process when count is more than one If CPU utilization is used, a high-low policy will need to be used CS 704D Advanced OS 28
- 29. Location Policies Location policies is one of the two following types Sender initiated Receiver initiated Depending on who initiates the search for a node that can accept the process CS 704D Advanced OS 29
- 30. Location-Sender Initiated When a node exceeds the threshold, it start broadcasting its status to find a node or starts random probes to other nodes to find a free node A probe limit is specified to stop the nodes from searching a large number of nodes and wasting time Scalability is better with limited probing Some analyses indicate the probe limit does not have much of an effect CS 704D Advanced OS 30
- 31. Location-Receiver Initiated A node load falls below threshold, it’ll start broadcast/probing to find a node that can provide a process A node can transfer a process only if doing that would not cause the load to go below the threshold With broadcast, the receiver can find one candidate from among the replies In random probing the probing continues until a candidate is found or the probing limit is reached CS 704D Advanced OS 31
- 32. Observations on Location Policies Either location policies offer advantage over no location of processes Sender initiated policies are preferred when the system is lightly or moderately loaded In a lightly loaded system, receiver generates probe messages would be many Receiver initiated policies are preferable on high load systems only if process transfer costs are comparable. As sender probes increase when the system is already highly loaded If transfer costs in receiver initiated policies are significantly higher then, sender initiated policies provide better performance CS 704D Advanced OS 32
- 33. State Information Exchange Policies No need to periodically exchange information State information is needed only when a node is under-loaded or overloaded Information is exchange only when state changes Policies used Broadcast when state changes Poll when state changes CS 704D Advanced OS 33
- 34. State Info-broadcast when State Changes When under-loaded or overloaded a node broadcasts a request for status information In sender initiated policy, the broadcast happens when the node is underloaded In receiver initiated policy the broadcast would be initiated when a node is changes state to underloaded When threshold is 1, the policy is called broadcast when idle CS 704D Advanced OS 34
- 35. State Info-Poll when State Changes When a node falls below the threshold, it initiates a random polling cycle The cycle ends when a node is found that can transfer a process to the receiver Or when a probing limit is reached When a fixed threshold of 1 is used, this is known as the poll when idle policy CS 704D Advanced OS 35