CS655_HessahAlsaaran_ProcessMigration

7 8 transfer strategies alternatives eager all

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: d node requests the migration. Ê༌  Trade- off between frequency and stability. Ê༌  Load Management and Migration Decisions Ê༌  Centralized or distributed. Ê༌  Suitable for lightly loaded systems Ê༌  Symmetric policy: combination of both of the above policies. Ê༌  Random policy: randomly choose a node. Ê༌  Simple but improves performance. (It can eliminate the management overhead, but it can miss the critical nodes that are overloaded or under loaded à༎ performance will depend on the load distribution.) 7 8 Transfer Strategies Alternatives Ê༌  Eager (all) strategy: Copy all address space at migration time. Ê༌  Remote Execution Ê༌  Eager (dirty) strategy: Transfer the dirty pages at migration time, then fault in the remaining when needed. Ê༌  Copy- on- reference strategy: Transfer pages on demand. If it is dirty then its from the source node. Otherwise, it can be from the source node or the backing store. Ê༌  Remote invocation of existing code, or transfer code only. Ê༌  Faster than process migration (no state transfer). Ê༌  But it doesn’t move. Ê༌  Once the the process is created at a node, it stays there until it terminates or get killed. Ê༌  Flushing strategy: same as above but the dirty pages are flushed to Ê༌  A process cannot resume execution; it is killed then restarted on the Ê༌  Precopy strategy: The process suspension at the source node is Ê༌  Can be suitable for very short processes that are not worth the backing store. other node. (So is it really an alternative?) delayed until transfer...
View Full Document

Ask a homework question - tutors are online