Nasf-Fan
Nasf-Fan
Only for test. Signed-off-by: Fan Yong
If a large transaction has sub-requests to dispatch to a lot of DTX participants, then we may have to split the dispatch process to multiple steps; otherwise, the dispatch process...
That will be helpful for log collection. Signed-off-by: Fan Yong
It is possible that some DTX on non-leader is blocked by long time bulk data transfer or DTX refresh, as to related RPC from DTX leader to non-leader for current...
When modify single replicated object, if it is not conditional, then it is unnecessary to keep related committed DTX entry, because: 1. We do not have DTX resync for such...
During DTX resync, if non-leader is excluded, handle it as 'prepared'. If other non-leaders are also 'prepared' then related DTX maybe still committable or 'corrupted'. The subsequent DTX resync logic...
For update request, it is possible that daos_obj_rw_t::sgls is NULL but daos_obj_rw_t::nr is non-zero, such as for punch_extent(). Such update request can be part of a distributed transaction. So related...
Only for test. Signed-off-by: Fan Yong ### Before requesting gatekeeper: * [ ] Two review approvals and any prior change requests have been resolved. * [ ] Testing is complete...
The patch fixes the following DTX related issues: 1. Use the address of tree handle for dtx_rpc_post. If we directly use the value of tree handle as dtx_rpc_post()'s input parameter,...
That will be helpful for log collection. Signed-off-by: Fan Yong