Hi qemu-devel community,

A gentle reminder and request for all migration maintainers - Peter, Juan, Dr. Gilbert and others too for review of the patchset series. Received reviewed-by from Daniel on migration implementation patches but need final approval from migration maintainers before getting it merged. Also got acked-by tag from Markus on the QAPI patches. This is Part1 of the 4 patchset series. Ultimate goal of the whole 4 series is to 'introduce multiple interface support on top of existing multifd capability'. Hope to get approval or comments from migration maintainers on the patches soon.

Thankyou :)

On 03/08/23 11:13 am, Het Gala wrote:
Hi,

A gentle reminder for Juan and other migration maintainers for the review of this patchset series if any changes are required or give to queue them. There are more patchset series coming after this. As discussed earlier, we have broken down it into 4 different patchset series. This is just Part1 of the 4 patchset series. Ultimate goal is to 'introduce multiple interface support on top of existing multifd capability'.

On 27/07/23 4:59 pm, Het Gala wrote:
This is just a ping for Juan and other migration maintainers, if it's possible to have a look at the migration patches for new QAPI design and suggest some review comments if any.

Update till now : Have got acked-by label from Markus for the new migrate QAPI design, and reviewd-by label from Daniel on the QAPI implementation side patches.

On 26/07/23 7:48 pm, Het Gala wrote:
This is v10 patchset of modified 'migrate' and 'migrate-incoming' QAPI design
for upstream review.

Would like to thank all the maintainers that actively participated in the v9 patchset discussion and gave insightful suggestions to improve the patches.


Link to previous upstream community patchset links:
v1: https://lists.gnu.org/archive/html/qemu-devel/2022-12/msg04339.html
v2: https://lists.gnu.org/archive/html/qemu-devel/2023-02/msg02106.html
v3: https://lists.gnu.org/archive/html/qemu-devel/2023-02/msg02473.html
v4: https://lists.gnu.org/archive/html/qemu-devel/2023-05/msg03064.html
v5: https://lists.gnu.org/archive/html/qemu-devel/2023-05/msg04845.html
v6: https://lists.gnu.org/archive/html/qemu-devel/2023-06/msg01251.html
v7: https://lists.gnu.org/archive/html/qemu-devel/2023-07/msg02027.html
v8: https://lists.gnu.org/archive/html/qemu-devel/2023-07/msg02770.html
v9: https://lists.gnu.org/archive/html/qemu-devel/2023-07/msg04216.html

v9 -> v10 changelog:
-------------------
- Patch6 : Added extra checks for migration arguments.
- Patch8 : Added checks for 'uri' and 'channels' both not present.
- Patch9 : Missed adding hmp_handle_error call to print error messages.
Abstract:
---------

Current QAPI 'migrate' command design (for initiating a migration
stream) contains information regarding different migrate transport mechanism (tcp / unix / exec), dest-host IP address, and binding port number in form of a string. Thus the design does seem to have some design issues. Some of the
issues, stated below are:

1. Use of string URIs is a data encoding scheme within a data encoding scheme.     QEMU code should directly be able to work with the results from QAPI,     without resorting to do a second level of parsing (eg. socket_parse()). 2. For features / parameters related to migration, the migration tunables needs     to be defined and updated upfront. For example, 'migrate-set-capability'     and 'migrate-set-parameter' is required to enable multifd capability and     multifd-number of channels respectively. Instead, 'Multifd-channels' can     directly be represented as a single additional parameter to 'migrate'     QAPI. 'migrate-set-capability' and 'migrate-set-parameter' commands could     be used for runtime tunables that need setting after migration has already
    started.

The current patchset focuses on solving the first problem of multi-level encoding of URIs. The patch defines 'migrate' command as a QAPI discriminated union for the various transport backends (like socket, exec and rdma), and on basis of transport backends, different migration parameters are defined.

(uri) string -->  (channel) Channel-type
                             Transport-type
                             Migration parameters based on transport type ------------------------------------------------------------------------------

Het Gala (10):
   migration: New QAPI type 'MigrateAddress'
   migration: convert migration 'uri' into 'MigrateAddress'
   migration: convert socket backend to accept MigrateAddress
   migration: convert rdma backend to accept MigrateAddress
   migration: convert exec backend to accept MigrateAddress.
   migration: New migrate and migrate-incoming argument 'channels'
   migration: modify migration_channels_and_uri_compatible() for new QAPI
     syntax
   migration: Implement MigrateChannelList to qmp migration flow.
   migration: Implement MigrateChannelList to hmp migration flow.
   migration: modify test_multifd_tcp_none() to use new QAPI syntax.

  migration/exec.c               |  72 +++++++++----
  migration/exec.h               |   8 +-
  migration/migration-hmp-cmds.c |  17 ++-
  migration/migration.c          | 190 ++++++++++++++++++++++++++-------
  migration/migration.h          |   3 +-
  migration/rdma.c               |  34 +++---
  migration/rdma.h               |   6 +-
  migration/socket.c             |  39 ++-----
  migration/socket.h             |   7 +-
  qapi/migration.json            | 150 +++++++++++++++++++++++++-
  softmmu/vl.c                   |   2 +-
  tests/qtest/migration-test.c   |   7 +-
  12 files changed, 409 insertions(+), 126 deletions(-)
Regards,
Het Gala
Regards,
Het Gala
Regards,
Het Gala

Reply via email to