Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | unify java warnings | Arne H Juul | 2022-01-06 | 1 | -7/+0 |
| | | | | | * these were stricter than in parent, but to simplify we can just use compiler args from parent | ||||
* | Add noexcept specifiers. | Tor Egge | 2021-12-11 | 3 | -3/+3 |
| | |||||
* | Update 2019 Oath copyrights. | gjoranv | 2021-10-27 | 1 | -1/+1 |
| | |||||
* | Update 2018 copyright notices. | gjoranv | 2021-10-07 | 3 | -3/+3 |
| | |||||
* | Update 2017 copyright notices. | gjoranv | 2021-10-07 | 81 | -81/+81 |
| | |||||
* | Update comments and use better naming. | Henning Baldersheim | 2021-09-23 | 2 | -17/+17 |
| | |||||
* | As NodeState is mutable a new one must be created when handing it out | Henning Baldersheim | 2021-09-23 | 1 | -2/+2 |
| | |||||
* | Description only matters if it is down.... | Henning Baldersheim | 2021-09-23 | 1 | -1/+1 |
| | |||||
* | Use a BitSet to enable more compact representation of ClusterState internally. | Henning Baldersheim | 2021-09-23 | 3 | -64/+98 |
| | | | | | | Down nodes without a comment occupies only a single bit. Up nodes that has no extra information also only occupies a single bit. The anomalities are represented in a hash map. | ||||
* | Refactor cluster state to avoid checking NodeType everywhere and reduce code ↵ | Henning Baldersheim | 2021-09-22 | 1 | -117/+166 |
| | | | | duplication. | ||||
* | Include cassert when needed. | Tor Egge | 2021-06-04 | 1 | -0/+1 |
| | |||||
* | Revert "Revert "GC unused DiskState and add the partition metrics to node ↵ | Henning Baldersheim | 2021-03-12 | 2 | -251/+0 |
| | | | | level."" | ||||
* | Revert "GC unused DiskState and add the partition metrics to node level." | Harald Musum | 2021-03-12 | 2 | -0/+251 |
| | |||||
* | GC unused DiskState and add the partition metrics to node level. | Henning Baldersheim | 2021-03-12 | 2 | -251/+0 |
| | |||||
* | Merge pull request #16926 from ↵ | Tor Brede Vekterli | 2021-03-12 | 2 | -21/+3 |
|\ | | | | | | | | | vespa-engine/vekterli/dont-store-full-bundle-objects-in-state-history Don't store full bundle objects in state history | ||||
| * | Don't store full bundle objects in cluster state history | Tor Brede Vekterli | 2021-03-12 | 2 | -21/+3 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | Bundles have a lot of sub-objects per state, so in systems with a high amount of node entries, this adds unnecessary pressure on the heap. Instead, store the string representations of the bundle and the string representation of the diff to the previous state version (if any). This is also inherently faster than computing the diffs on-demand on every status page render. Also remove mutable `official` field from `ClusterState`. Not worth violating immutability of an object just to get some prettier (but with high likelihood actually more confusing) status page rendering. | ||||
* | | Revert "GC unused DiskState" | Arnstein Ressem | 2021-03-12 | 2 | -0/+251 |
|/ | |||||
* | GC unused DiskState | Henning Baldersheim | 2021-03-11 | 2 | -251/+0 |
| | |||||
* | GC unused diskstates array. | Henning Baldersheim | 2021-03-11 | 1 | -142/+5 |
| | |||||
* | GC use of void DiskState. | Henning Baldersheim | 2021-03-11 | 3 | -63/+18 |
| | |||||
* | GC use of NodeState.getDiskCount and NodeState.getDiskStates. | Henning Baldersheim | 2021-03-11 | 4 | -106/+2 |
| | |||||
* | GC long gone disk state checks. | Henning Baldersheim | 2021-03-11 | 2 | -24/+0 |
| | |||||
* | Shrink the size of the NodeState object by using float over double for ↵ | Henning Baldersheim | 2021-03-11 | 3 | -14/+13 |
| | | | | initProgress and capacity. Also gc unused 'reliability' member. | ||||
* | Remove notion of node-specific reliability from Java distribution code | Tor Brede Vekterli | 2021-03-02 | 4 | -62/+24 |
| | | | | | | | | Not used, and wasn't algorithmically in sync with the C++ code anyway. Also add guard to avoid emitting invalid node indices for storage nodes if the number of configured nodes is lower than the replication factor. Looks like this particular code path is only called by cross-language conformance tests, so hasn't been a problem in practice. | ||||
* | Remove notion of node-specific reliability from C++ distribution code | Tor Brede Vekterli | 2021-03-02 | 5 | -119/+32 |
| | | | | | I have never seen this in use anywhere, and can find no code that ever sets it. Bonus is that node candidate trimming can be vastly simplified. | ||||
* | string.h does not need alloc.h | Henning Baldersheim | 2021-02-21 | 4 | -0/+4 |
| | |||||
* | - Reduce visibility of ClusterState and Distribution. | Henning Baldersheim | 2021-02-19 | 9 | -51/+27 |
| | |||||
* | use size literals in streaming | Arne Juul | 2021-02-15 | 3 | -5/+8 |
| | |||||
* | Revert "Revert "Jonmv/reapply document protocol super config [run-systemtest]"" | Jon Marius Venstad | 2021-01-28 | 4 | -66/+122 |
| | | | | This reverts commit 2e2e2edeb3ea99f2c04925070cf44601e2cd94fb. | ||||
* | Revert "Jonmv/reapply document protocol super config [run-systemtest]" | Jon Marius Venstad | 2021-01-28 | 4 | -122/+66 |
| | |||||
* | Revert "Revert "Jonmv/document protocol super config"" | Jon Marius Venstad | 2021-01-27 | 4 | -66/+122 |
| | | | | This reverts commit d7359f7c72ff06889af594431baf4075e2b4da78. | ||||
* | Revert "Jonmv/document protocol super config" | Harald Musum | 2021-01-27 | 4 | -122/+66 |
| | |||||
* | Merge pull request #16089 from vespa-engine/jonmv/document-protocol-super-config | Jon Marius Venstad | 2021-01-27 | 4 | -66/+122 |
|\ | | | | | Jonmv/document protocol super config | ||||
| * | Address review | Jon Marius Venstad | 2021-01-27 | 1 | -3/+3 |
| | | |||||
| * | Use direct bit operations instead of stored array | Jon Marius Venstad | 2021-01-19 | 2 | -17/+12 |
| | | |||||
| * | Use only injected config for injected document access | Jon Marius Venstad | 2021-01-19 | 2 | -47/+97 |
| | | |||||
| * | Add config for distribution and doc-protoc policies across _all_ content ↵ | Jon Marius Venstad | 2021-01-18 | 2 | -4/+15 |
| | | | | | | | | clusters | ||||
* | | Merge pull request #16103 from vespa-engine/hakonhall/support-group-maintenance | Håkon Hallingstad | 2021-01-20 | 1 | -0/+2 |
|\ \ | | | | | | | Support group maintenance [run-systemtest] | ||||
| * | | Support group maintenance [run-systemtest] | Håkon Hallingstad | 2021-01-19 | 1 | -0/+2 |
| |/ | |||||
* / | Add feed block status to ClusterStateBundle in C++. | Geir Storli | 2021-01-18 | 3 | -6/+111 |
|/ | |||||
* | Merge pull request #16014 from vespa-engine/balder/disable-valgrind | Henning Baldersheim | 2021-01-13 | 1 | -5/+1 |
|\ | | | | | Reduce cost as test from 400 to below 100 | ||||
| * | Update CMakeLists.txt | Henning Baldersheim | 2021-01-13 | 1 | -1/+1 |
| | | |||||
| * | No need to set cost explicit. | Henning Baldersheim | 2021-01-12 | 1 | -1/+1 |
| | | |||||
| * | Disable valgrind as this test takes 360s with and 3s without. | Henning Baldersheim | 2021-01-12 | 1 | -5/+1 |
| | | |||||
* | | - Factor out insert() to a separate method. | Henning Baldersheim | 2021-01-12 | 10 | -122/+103 |
| | | | | | | | | | | | | | | | | | | - Inline Node::Node. - Remove virtuality of Node - Inline State::oneOf - Microoptimize ClusterState::getNodeState. This brings the runtime down from 2.23s to 1.68s And with valgrind from 130s to 85s. In total with the other commits in this PR it is now down from 2.95s to 1.68, and from 360s to 85s with valgrind. | ||||
* | | Allocate once and reuse. | Henning Baldersheim | 2021-01-12 | 1 | -21/+22 |
| | | |||||
* | | Replace the use of std::list with std::vector. This avoids a lof of new/delete. | Henning Baldersheim | 2021-01-12 | 1 | -28/+16 |
|/ | | | | | The number of elements that are moved around is small and significantly less than the gain by more efficient memory management. | ||||
* | Eliminate inlining warnings. | Tor Egge | 2020-12-04 | 2 | -0/+4 |
| | |||||
* | Add noexcept | Henning Baldersheim | 2020-11-26 | 2 | -11/+11 |
| | |||||
* | As we have have now removed the expensive Route member we can further ↵ | Henning Baldersheim | 2020-11-26 | 2 | -20/+36 |
| | | | | | | | | | | compact the message objects. - Compact StorageMessageAddress to 16 bytes by - using reference to cluster name. - Use small enums for protocol and node type. - Avoid having StorageMessage as separate allocation. - Avoid default values |