Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: I4e9cd08ad8640b2733fa1c62cbd620a639e81bed
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Introduces YangDAG that maintains topological order for yang schema
nodes and can be used by other components to read additional graph edges,
e.g. from file (HONEYCOMB-365).
Change-Id: Ia3046d38ffb4ca222412309f6c4391afc9315bd2
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Only ReaderRegistry and WriterRegistry instances are needed by other components.
Change-Id: Icc026154e80be4ca9424c79754f342e93fda0735
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I1a98ea27dde5ba77f8b20382dfeb29496f3282b2
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Whitout specifying it explicitly Guice might create more instances
Change-Id: Ia3cf56e54d59fa49488ab93a26a9db44cec51e8d
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Given current YangBindingProviderModule implementation (HONEYCOMB-358),
needed to preserve footprint of minimal distro.
The change also:
- reverts logger&persistence config changes
introduced by https://gerrit.fd.io/r/#/c/5972/
- improves logging for BGP component initialization
Change-Id: Ie223833518be21628edb7ace36c7618e427585e0
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Features:
- static bgp instance
- static peer registry (both remote & app peers supported)
- bgp operational state exposed via RESTCONF/NETCONF
- bgp-inet support
- postman collection with examples
TODOs:
- make bgp disabled by default in honeycomb.json (probably bgp should be build using profile
or allow to filter yangs loaded by SchemaCtx, otherwise footprint will grow)
- update asciidocs for infra module
- SPI for VPP integration:
- allow RIB changes listeners
- or writers (that can reject changes) to mirror netconf/restconf design
- expose bgp server & peer cfg via netconf/restconf (currently read from json files)
- test support for all BGP extensions (flowspec BGP needs additional care)
- make bgp extensions fully configurable (probably requires changes in ODL)
Testing
1) start hc distro:
./infra/minimal-distribution/target/minimal-distribution-1.17.07-SNAPSHOT-hc/minimal-distribution-1.17.07-SNAPSHOT/honeycomb
2) run bgp-testtool:
java -jar ./target/bgp-testtool-0.6.3-Boron-SR3-executable.jar
-ra 127.0.0.1:1790 -la 127.0.0.2:17900 -pr 2 -ll DEBUG --active true -as 65000 -mp true
then GET http://localhost:8183/restconf/operational/bgp-rib:bgp-rib
Change-Id: I156d6a1c79174279db8c8936c40ac1f28e14794c
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Removes unsupported AbstractDataTreeConverter
Change-Id: I0b45f1d16139a55f2248294b57bfad6d8ee32c45
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I6bc2f99806f81c206dcf31711a01fcebe809288a
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Allows these types of modifications to check in depth,
to see if some of their children nodes were not modified
Change-Id: Ice2f988732c2d9ecad8e960c4f10d01863fb0cfd
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: I644c229ce82e4ea9feb9d53a6fe059a7fbe15d77
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: Ibd21e21932ddaff514221c0e5f111df3120e4d8e
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
(cherry picked from commit 192d742f43a0272bfe73613630ccbfc9cb5d09e0)
|
|
Boron SR3 contains fixes for HC2VPP-62,HC2VPP-69,HC2VPP-68
Change-Id: I06ad40da6fe2f702c735bfa5413de19ba3132046
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
All ODL dependencies are in one place now.
Change-Id: I0fc91f7efcf81e002dcbef9acfef1d82eac80238
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: Ie7b486c266ea34df999e7122b5ef12d26aded863
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I72f0e5339d9fc39a3ca76c67a20e8358ce7f700a
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I18b2ea3c897c467740f19bf346d13240aac458ac
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: Ief2224325fb7b9ec1dd3bbb21536f6306b68aa1a
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: Ia02ed73daaeea547d49c9ec2d4d7d10f4db85b5e
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
checks if provided resource path was found
Change-Id: I86ed1b6f38bc8f48fd12ebe8620831862840ed11
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Workaround for HC2VPP-62
until
https://bugs.opendaylight.org/show_bug.cgi?id=7759
is fixed.
Change-Id: I0e9c82378865bf99c1843b0399940a53350e2d76
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I4585cbc2100ef0a76e50e870d57b28221debf2f6
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: I96e7db8f295c9c3d5b14395c7785574d12d76ea9
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Major changes
- mechanism to inject list entries by key
- provided processor registry to hide explicit implementations
Minor changes
- general refactoring
Test cases
- list in root of model
- list under container
- list under nested container
- list in augmentation
Change-Id: I9abe1ce5f9176c132ad88627b135516574e40e06
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Static customizer must be provided instead till
HONEYCOMB-344 is resolved
Change-Id: Id71f69fe23b156e6de10eed06a3596de8e602bbe
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Contains configuration for benchmarks:
- Honeycomb on rest(just infra, no data)
- Honeycomb with 1000 data nodes
- Honeycomb with 10000 data nodes
Each benchmark outputs two files:
- out_path_name-HeapMemoryUsage.csv
- out_path_name-NonHeapMemoryUsage.csv
Both files are in format :
committed,init,max,used
109576192,109051904,1525153792,12194752
Data sample sizes can be easily adjusted, just by changing
start parameter -DsampleSize
Change-Id: If6f9919307574237689326b4a38d410ec563200a
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: I60f1b3f79ddb578d6fca157fe5736de40b30623e
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
(cherry picked from commit 78886acd688284585c2e219e18d7289f49cc8a45)
|
|
Unable to achieve same results with maven resource plugin.
Change-Id: I564f00dcabca638b4bee40fe05ff8d4505623832
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Due to incorrect instance of check, InitSubtreeReader was never created
in CompositeReaderRegistryBuilder.getSubtreeHandler().
As a consequence, initializers registered by subtreeAdd*, were not invoked.
The patch:
- fixes mentioned check,
- makes readers implement InitReader instead of directly
implementing Initializer
- includes missing unit tests for GenericReaders.
Change-Id: I93be59fafddb60dce00191958b5c8c62e7c2d289
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I99bc2593d50e59a6d4293e2d98ef3a9569a55008
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I187b6b6b1c4c2812a24b4aad7cf9bc40d0f0de0c
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Fixes HONEYCOMB-290 among others
Change-Id: I3902dacc2cc8279abec43cd3b4d74074485cf7e0
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Fixes Status.OK on fail with successful revert (HONEYCOMB-297)
Change-Id: I3862347775eea08a62671d5755f95b07f64e99bb
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
plugin
Generates 3 plots for write operations:
- simple-container.csv,
- list-in-container.csv,
- complex-list-in-container.csv
and one plot for read operations: operational-read.csv
from benchmark.csv.
Jenkins plot plugin uses only first record line from CSV file:
https://github.com/jenkinsci/plot-plugin/blob/master/src/main/webapp/help-csv.html#L15
therefore each generated CSV contains two lines: header and single record.
Change-Id: I489c3401673d72a4b3ea3ec00197366118d1ff5f
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: Iccffe5412b4bb06b606b66f7c0e7ebd601d5a7d1
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I47f37a60f276928672fc2bd53e948641d9278014
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Standard cache key factory made type-aware
Added checking for type of returned data from cache
Change-Id: Ie4d31a9d2b0d25c4b2f4ea66be98060f449007b6
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: Ib7bcde546faf8e9bf73d16adaf87c2f8c43ec4aa
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: I6b24a1d061c94d57730b20cde4c5b3c39444b119
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Moved to a dedicated hc2vpp project in fd.io
Change-Id: I03dc3b3029f21b127a00c69a86bcd8e467896241
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: Id887c12bea1983c0d185689c633902b691cf52f6
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: Ia086f3b677fe33ecb20a3ded780dac9f89415590
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Added mapping context between adjacency id(HC side)
and combination of remote and local eids(vpp side)
Change-Id: Icbba9130a82b7af51b23e6d6d68fcfbf49e66c99
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: Iaf6c5ce430dd793a9cfff5527b910a79eab9aa2e
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: I7c57d7c525bf29dde922d48c1c8b8b72cdec9980
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: Ia0986dbe173123a4dca4f7f3bf65e7ba4851b1b7
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
So that they can influence whether empty data is to be considered as present
+ Move registries implementations from util to impl
+ Introduce DelegatingReader trait
+ Extend GenericReader where possible to reduce duplication
Change-Id: I5a416acd0c4eab1fbc30fcbe585719991dbe9215
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Config (write)
Operational (read)
Add -Pbenchmark to maven execution to include benchmarks
Change-Id: Ia4815ffc109e34629279b9418b962a9f91c38c30
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: I9db2d9705abeb27be0fb0c71229386b1943dcaf9
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Added mandatory for sub-tables under vni-table
Change-Id: I0f410cc0e79becafc16e7012395fa6269b278b98
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|