Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Use std::max, and do it on the correct place. | Henning Baldersheim | 2018-06-15 | 1 | -1/+1 |
| | |||||
* | Control threadpool and dispatch of encode/decode by config. | Henning Baldersheim | 2018-06-15 | 2 | -4/+20 |
| | |||||
* | 1 - Use a backing buffer for the DocumentUpdate that always is source of truth. | Henning Baldersheim | 2018-06-12 | 9 | -39/+25 |
| | | | | | | | | | | | 2 - Use this buffer for re-serialization. 3 - Make deserialization lazy where possible. Currently lazy on replay and when arriving over the storageapi. Still needs to eager over documentapi. 4 - Deserialize eagerly in the persistence thread since those are plentyfull and not bottlenecked, instead of in the single master thread. Use real repo. | ||||
* | Last style changes before lazy documentupdate PR. | Henning Baldersheim | 2018-06-12 | 3 | -37/+13 |
| | |||||
* | Merge branch 'master' into balder/reduce-code-visibility-rebased | Henning Baldersheim | 2018-06-12 | 2 | -8/+0 |
|\ | |||||
| * | clean up remnants of .mak files | Arne Juul | 2018-06-11 | 2 | -8/+0 |
| | | |||||
* | | Reduce code visibility | Henning Baldersheim | 2018-06-12 | 8 | -179/+90 |
|/ | |||||
* | Don't assume visitor threads immediately destroy visitors | Tor Brede Vekterli | 2018-06-01 | 1 | -11/+7 |
| | |||||
* | Clean up some old code | Tor Brede Vekterli | 2018-06-01 | 1 | -79/+44 |
| | |||||
* | Clean up visitors that we've already initiated | Tor Brede Vekterli | 2018-06-01 | 1 | -10/+37 |
| | |||||
* | Simplify tests | Tor Brede Vekterli | 2018-06-01 | 1 | -45/+2 |
| | |||||
* | Wait for all started visitors to finish | Tor Brede Vekterli | 2018-06-01 | 1 | -14/+30 |
| | | | | Also correct wrong expected number of busy bounces in test | ||||
* | Revert "Ensure visitormanager tests clean up after themselves" | Henning Baldersheim | 2018-05-31 | 1 | -113/+148 |
| | |||||
* | Merge pull request #6035 from ↵ | Henning Baldersheim | 2018-05-31 | 1 | -148/+113 |
|\ | | | | | | | | | vespa-engine/vekterli/ensure-visitormanager-tests-clean-up-after-themselves Ensure visitormanager tests clean up after themselves | ||||
| * | Clean up some old code | Tor Brede Vekterli | 2018-05-31 | 1 | -79/+44 |
| | | |||||
| * | Clean up visitors that we've already initiated | Tor Brede Vekterli | 2018-05-31 | 1 | -10/+37 |
| | | |||||
| * | Simplify tests | Tor Brede Vekterli | 2018-05-31 | 1 | -45/+2 |
| | | |||||
| * | Wait for all started visitors to finish | Tor Brede Vekterli | 2018-05-31 | 1 | -14/+30 |
| | | | | | | | | Also correct wrong expected number of busy bounces in test | ||||
* | | Implicitly enable multiple bucket spaces in content backend | Tor Brede Vekterli | 2018-05-31 | 1 | -7/+3 |
|/ | | | | | Config can still force new protocol version to be used even if version handshake indicates it should not be. | ||||
* | Remap missing bucket space for DocumentAPI Gets to "Not Found" | Tor Brede Vekterli | 2018-05-30 | 3 | -9/+51 |
| | | | | | | Lets legacy routing of Gets to all clusters successfully merge responses even if a recipient cluster does not have a valid mapping for the requested bucket space. | ||||
* | Use fallthrough attributes. | Tor Egge | 2018-05-27 | 1 | -2/+2 |
| | |||||
* | Disperse bucket bits using FNV-1 prime when choosing thread stripe | Tor Brede Vekterli | 2018-05-23 | 1 | -1/+7 |
| | | | | | This avoids an inherent affinity between the LSB of a bucket's bits and the stripe an operation ends up on. | ||||
* | Catch by reference | Henning Baldersheim | 2018-05-11 | 1 | -2/+2 |
| | |||||
* | Remove redundant typedef | Henning Baldersheim | 2018-05-11 | 1 | -7/+1 |
| | |||||
* | Ensure commands are processed before shutting down persistence threads | Tor Brede Vekterli | 2018-04-27 | 2 | -24/+28 |
| | | | | | | | Prevents race condition where test cleanup closes persistence threads while there are still enqueued operations. Normally the content layer shall ensure that all enqueued operations are aborted before reaching this step, but this does not take place in the simplified testing setup. | ||||
* | Make test names less ambiguous | Tor Brede Vekterli | 2018-04-26 | 1 | -6/+6 |
| | |||||
* | Don't be silly and deref an iterator after it has been erased | Tor Brede Vekterli | 2018-04-26 | 1 | -1/+1 |
| | |||||
* | Handle explicit node state response case without pending request | Tor Brede Vekterli | 2018-04-26 | 3 | -19/+105 |
| | | | | | | | | | | | | | Deals with the edge case where a node explicitly triggers a node state response to be sent (generally due to updated host info to the controller(s)) but where there is no pending request to immediately reply to. We now maintain state per controller which lets us infer whether it needs to immediately observe a reply when the next request is received for it. This avoids having to wait for an entire RPC timeout cycle before a response is sent to the controller. This feature is predominantly to avoid keeping nodes in the default bucket space in implicit maintenance mode for longer than is necessary when merges in the global space are completed. | ||||
* | Make rpc target cache ttl configurable with a deafult lifted from 30s to 10 ↵ | Henning Baldersheim | 2018-04-18 | 2 | -2/+6 |
| | | | | minutes. | ||||
* | Remove no longer needed test | Henning Baldersheim | 2018-04-16 | 1 | -25/+0 |
| | |||||
* | Remove no longer needed hash normalizer. | Henning Baldersheim | 2018-04-16 | 8 | -366/+5 |
| | |||||
* | Remove strange temporary value pointer code | Tor Brede Vekterli | 2018-04-13 | 2 | -14/+9 |
| | | | | | | | Overriding unary & is potentially dangerous and the code does not give any justification for why it is done. Nor should it be necessary, since passing an rvalue as a const ref into the LoadMetric to be used as a template should suffice plenty. | ||||
* | Tidy up some existing code | Tor Brede Vekterli | 2018-04-13 | 2 | -27/+25 |
| | |||||
* | Add request size metric for Put, Update, Remove and Get commands | Tor Brede Vekterli | 2018-04-13 | 4 | -13/+138 |
| | | | | Request size is approximated by the network payload size for the commands. | ||||
* | Use 2 stripes | Henning Baldersheim | 2018-04-12 | 1 | -3/+1 |
| | |||||
* | Add and use getTotal | Henning Baldersheim | 2018-04-11 | 1 | -1/+1 |
| | |||||
* | Add back the original avgerage queue wait time metric | Henning Baldersheim | 2018-04-10 | 3 | -5/+18 |
| | |||||
* | Unsigned...... | Henning Baldersheim | 2018-04-10 | 1 | -1/+1 |
| | |||||
* | Start with only a single stripe. | Henning Baldersheim | 2018-04-10 | 1 | -1/+1 |
| | |||||
* | Do not pregenerate a summary string from message that is 'never' needed. | Henning Baldersheim | 2018-04-10 | 2 | -14/+17 |
| | |||||
* | Move averargeQ time to Stripe. | Henning Baldersheim | 2018-04-10 | 10 | -60/+92 |
| | |||||
* | Rename member to reflect sematic meaning of stripId, not threadId | Henning Baldersheim | 2018-04-09 | 2 | -5/+5 |
| | |||||
* | Consistent parameter naming in definition and implementation. | Henning Baldersheim | 2018-04-09 | 1 | -6/+6 |
| | |||||
* | remove comment bringing now value to the table | Henning Baldersheim | 2018-04-09 | 1 | -5/+1 |
| | |||||
* | Use stripeId | Henning Baldersheim | 2018-04-09 | 1 | -1/+1 |
| | |||||
* | One thread per strip | Henning Baldersheim | 2018-04-09 | 1 | -1/+1 |
| | |||||
* | Remove assert that is invalid as threads are started upon construction | Henning Baldersheim | 2018-04-09 | 1 | -1/+0 |
| | |||||
* | No need to const cast | Henning Baldersheim | 2018-04-09 | 1 | -2/+2 |
| | |||||
* | Remove unused includes. | Henning Baldersheim | 2018-04-09 | 2 | -6/+2 |
| | |||||
* | Start with a single stripe for now. | Henning Baldersheim | 2018-04-09 | 1 | -1/+1 |
| |