aboutsummaryrefslogtreecommitdiffstats
path: root/vdslib
Commit message (Collapse)AuthorAgeFilesLines
* switch to new-style configArne H Juul2022-04-121-91/+113
|
* Avoid requiring identifiable for FieldValueHenning Baldersheim2022-03-162-5/+4
|
* Add include to get declaration of std::sort (vdslib).Tor Egge2022-02-271-0/+1
|
* Stop using std::binary_function (vdslib).Tor Egge2022-02-262-2/+2
|
* Make ConfigUri constructors explicit and use same context where possible in ↵Henning Baldersheim2022-02-201-2/+1
| | | | proton.
* Allow inhibiting all default bucket space merges if global merges pendingTor Brede Vekterli2022-02-152-8/+8
| | | | | | | | | | | | | | | | | | If the cluster controller publishes a cluster state where at least one node is marked as maintenance in the default space but is up in the global space, global bucket merges are pending in the cluster. We want these merges to complete ASAP, and to avoid starvation caused by distributors that finish their global merges quicker than others (or whose global bucket ops get preempted), allow for _completely_ inhibiting merges to buckets in the default space when global buckets are pending. Requires a live config flag to be set to actually be taken into account in the bucket synchronization state checker. Flag is tracked per distributor bucket space (per stripe) and is automatically cleared once a cluster state is received without pending global merges.
* Reduce code visibility and include only what you need from config library.Henning Baldersheim2022-02-061-4/+1
|
* Merge pull request #21013 from vespa-engine/deprecate-config.subscriptionHarald Musum2022-02-022-3/+5
|\ | | | | Deprecate config.subscription
| * Add SuppressWarnings for removal of config.subscription types.gjoranv2022-02-021-0/+3
| |
| * Add TODO to not use ConfigGettergjoranv2022-01-311-3/+2
| |
* | Merge pull request #21001 from ↵Henning Baldersheim2022-01-311-0/+25
|\ \ | | | | | | | | | | | | vespa-engine/vekterli/deprecate-legacy-visitor-functionality Deprecate legacy visitor functionality
| * | Deprecate remnants of two-phase visiting APIs and usageTor Brede Vekterli2022-01-311-0/+25
| |/ | | | | | | | | | | | | | | | | | | This has not been relevant since "orderdoc" was a thing, and it was never really a thing in the first place. Unfortunately, due to an oversight in how the backend propagates visitor statistics, the internal usage of 2nd pass statistics cannot be fully removed before Vespa 8 (where the backend is known to not set the deprecated statistics fields).
* / Don't set deprecated fields when aggregating visitor statisticsTor Brede Vekterli2022-01-311-2/+2
|/ | | | | | | | The 1st/2nd pass functionality has been deprecated for a long time, but unfortunately the documents/bytes visited stats have been wired to be returned as part of 2nd phase statistics instead of the regular higher-level fields. This commit changes this, but the serialization will still have to remain in place until Vespa 8.
* unify java warningsArne H Juul2022-01-061-7/+0
| | | | | * these were stricter than in parent, but to simplify we can just use compiler args from parent
* Add noexcept specifiers.Tor Egge2021-12-113-3/+3
|
* Update 2019 Oath copyrights.gjoranv2021-10-271-1/+1
|
* Update 2018 copyright notices.gjoranv2021-10-073-3/+3
|
* Update 2017 copyright notices.gjoranv2021-10-0781-81/+81
|
* Update comments and use better naming.Henning Baldersheim2021-09-232-17/+17
|
* As NodeState is mutable a new one must be created when handing it outHenning Baldersheim2021-09-231-2/+2
|
* Description only matters if it is down....Henning Baldersheim2021-09-231-1/+1
|
* Use a BitSet to enable more compact representation of ClusterState internally.Henning Baldersheim2021-09-233-64/+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 Baldersheim2021-09-221-117/+166
| | | | duplication.
* Include cassert when needed.Tor Egge2021-06-041-0/+1
|
* Revert "Revert "GC unused DiskState and add the partition metrics to node ↵Henning Baldersheim2021-03-122-251/+0
| | | | level.""
* Revert "GC unused DiskState and add the partition metrics to node level."Harald Musum2021-03-122-0/+251
|
* GC unused DiskState and add the partition metrics to node level.Henning Baldersheim2021-03-122-251/+0
|
* Merge pull request #16926 from ↵Tor Brede Vekterli2021-03-122-21/+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 historyTor Brede Vekterli2021-03-122-21/+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 Ressem2021-03-122-0/+251
|/
* GC unused DiskStateHenning Baldersheim2021-03-112-251/+0
|
* GC unused diskstates array.Henning Baldersheim2021-03-111-142/+5
|
* GC use of void DiskState.Henning Baldersheim2021-03-113-63/+18
|
* GC use of NodeState.getDiskCount and NodeState.getDiskStates.Henning Baldersheim2021-03-114-106/+2
|
* GC long gone disk state checks.Henning Baldersheim2021-03-112-24/+0
|
* Shrink the size of the NodeState object by using float over double for ↵Henning Baldersheim2021-03-113-14/+13
| | | | initProgress and capacity. Also gc unused 'reliability' member.
* Remove notion of node-specific reliability from Java distribution codeTor Brede Vekterli2021-03-024-62/+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 codeTor Brede Vekterli2021-03-025-119/+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.hHenning Baldersheim2021-02-214-0/+4
|
* - Reduce visibility of ClusterState and Distribution.Henning Baldersheim2021-02-199-51/+27
|
* use size literals in streamingArne Juul2021-02-153-5/+8
|
* Revert "Revert "Jonmv/reapply document protocol super config [run-systemtest]""Jon Marius Venstad2021-01-284-66/+122
| | | | This reverts commit 2e2e2edeb3ea99f2c04925070cf44601e2cd94fb.
* Revert "Jonmv/reapply document protocol super config [run-systemtest]"Jon Marius Venstad2021-01-284-122/+66
|
* Revert "Revert "Jonmv/document protocol super config""Jon Marius Venstad2021-01-274-66/+122
| | | | This reverts commit d7359f7c72ff06889af594431baf4075e2b4da78.
* Revert "Jonmv/document protocol super config"Harald Musum2021-01-274-122/+66
|
* Merge pull request #16089 from vespa-engine/jonmv/document-protocol-super-configJon Marius Venstad2021-01-274-66/+122
|\ | | | | Jonmv/document protocol super config
| * Address reviewJon Marius Venstad2021-01-271-3/+3
| |
| * Use direct bit operations instead of stored arrayJon Marius Venstad2021-01-192-17/+12
| |
| * Use only injected config for injected document accessJon Marius Venstad2021-01-192-47/+97
| |
| * Add config for distribution and doc-protoc policies across _all_ content ↵Jon Marius Venstad2021-01-182-4/+15
| | | | | | | | clusters