From: "Dr. David Alan Gilbert" <dgilb...@redhat.com> Hi, Postcopy currently turns off bandwidth limits during the postcopy phase to make sure the urgent postcopy requests aren't delayed. This causes problems for larger clusters which share networking between the migration stream and other critical services.
This series restricts the background postcopy bandwidth but does it in a way that lets the urgent postcopy requests get through. Testing on a 10Gbps link and a 400MByte/s limit shows very little difference in the postcopy request latency, but a network graph shows the bandwidth usage very close to the set limit, even with a very heavy memory load running in the guest. Dr. David Alan Gilbert (3): migration/postcopy: Add max-postcopy-bandwidth parameter migration: Wake rate limiting for urgent requests migration/postcopy: Wake rate limit sleep on postcopy request hmp.c | 7 +++++ migration/migration.c | 70 +++++++++++++++++++++++++++++++++++++++--- migration/migration.h | 8 +++++ migration/ram.c | 9 +++++- migration/trace-events | 2 ++ qapi/migration.json | 19 ++++++++++-- 6 files changed, 106 insertions(+), 9 deletions(-) -- 2.17.1
