From 706b886bca8d7f51596da3cd0002c9e017137ade Mon Sep 17 00:00:00 2001 From: denis-tingaikin Date: Fri, 6 Oct 2023 21:17:24 +0300 Subject: [PATCH] run 'make fetch-notes manually' Signed-off-by: denis-tingaikin --- content/docs/releases/v1.11.0.md | 145 +++++++++++++++++++++++++------ 1 file changed, 119 insertions(+), 26 deletions(-) diff --git a/content/docs/releases/v1.11.0.md b/content/docs/releases/v1.11.0.md index d5fa8c4..cdebf29 100755 --- a/content/docs/releases/v1.11.0.md +++ b/content/docs/releases/v1.11.0.md @@ -6,14 +6,7 @@ date="2023-10-06 00:00:00 +0000 UTC" +++ -# Schedule - -Feature Freeze: 2023-09-19 \ -Release Candidate: 2023-09-26 \ -Release: 2023-10-06 - - -# [DRAFT] NSM Release v1.11.0 +# NSM Release v1.11.0 NSM v1.11.0 has been tested on: @@ -26,46 +19,81 @@ NSM v1.11.0 has been tested on: ## Changes since last release +### Add LoadBalancer for vl3 networks +NSM vl3 could be perceived as an alternative to k8s networks. Currently, k8s service is the most useful thing in the k8s network. -### Update vpp version +At this moment, NSM vl3 netowk doesn't provide any alternative to k8s services. So we need to consider and implement the best solution for it. -Status: RESOLVED. +[See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9210) -[See more details](https://github.com/networkservicemesh/govpp/issues/9) +### Add interdomain healing examples +Currently, we don't test healing over interdomain or floating interdomain. So we could add some examples of how to do that. -## System stability fixes and improvements +Scenarios: +- Forwarders death in floating interdomain scenario +- NSE death in floating interdomain scenario +- NSM systems death in floating interdomain scenario +- Proxy nsmgrs death in interdomain scenario +- NSMGRs death in interdomain scenario +- Registry death in interdomain scenario +We need to check scenarious from single cluster testing in multicluster scenarious. +[See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9647) -### Add VPP config for ARM - - -Add vpp config for ARM. The config will be the same as in this PR - https://github.com/networkservicemesh/cmd-forwarder-vpp/pull/905 +### Add health checks for vpp based apps +We'd added datapath checks for kernel ifaces. https://github.com/networkservicemesh/sdk-kernel/blob/main/pkg/kernel/tools/heal/liveness_check.go -[See more details](https://github.com/networkservicemesh/vpphelper/issues/1) +[See more details](https://github.com/networkservicemesh/sdk-vpp/issues/737) + + +## System stability fixes and improvements -### Update all cmd-*-vpp repos that use edwarnicke/vpphelper + +### No artifacts after CI run on public clusters Status: RESOLVED. -[See more details](https://github.com/networkservicemesh/vpphelper/issues/4) +[See more details](https://github.com/networkservicemesh/integration-tests/issues/2874) ### `Update dependent repositories` workflow does update incorrectly -`integration-k8s-kind` re-uses [.github workflow](https://github.com/networkservicemesh/.github/blob/main/.github/workflows/update-dependent-repositories-gomod.yaml) -This is wrong, because this workflow updates `integration-k8s-kind` instead of `integration-tests` in dependent repositories (public clusters). -For example: +`integration-k8s-kind` re-uses [.github workflow](https://github.com/networkservicemesh/.github/blob/main/.github/workflows/update-dependent-repositories-gomod.yaml) +This is wrong, because this workflow updates `integration-k8s-kind` instead of `integration-tests` in dependent repositories (public clusters). +For example: https://github.com/networkservicemesh/integration-k8s-kind/actions/runs/5656999510/job/15325140836 [See more details](https://github.com/networkservicemesh/integration-k8s-kind/issues/862) +### How can i make composition of endpoints? + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9676) + + +### IPSec takes a lot of time on request + +Our current implementation uses IKEv2, that requires `rsa` key. +The thing is the key generation takes about 3-5s - https://github.com/networkservicemesh/sdk-vpp/blob/main/pkg/networkservice/mechanisms/ipsec/common.go + +[See more details](https://github.com/networkservicemesh/sdk-vpp/issues/745) + + +### Routing between 2 containers + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9864) + + ### Feature Request: Allow configuration of initContainer resource requests and limits Status: RESOLVED. @@ -73,6 +101,20 @@ Status: RESOLVED. [See more details](https://github.com/networkservicemesh/cmd-admission-webhook-k8s/issues/292) +### MatchesMonitorScopeSelector helper function checks only one field + +[MatchesMonitorScopeSelector](https://github.com/networkservicemesh/api/blob/main/pkg/api/networkservice/connection_helpers.go + +[See more details](https://github.com/networkservicemesh/api/issues/164) + + +### Add release automation for integration-k8s-${platform} repos + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/integration-tests/issues/1667) + + ### Allow setting of registration url via env variable Status: RESOLVED. @@ -87,6 +129,23 @@ Status: RESOLVED. [See more details](https://github.com/networkservicemesh/.github/issues/40) +### Log rotated after 5 minutes in NSMgr / Registry + + + +To fix a possible increased latency due to logging in forwarder-vpp we could add a new chain element for min logging if current log level is not equal TRACE + +- Log input to server chain + - Return from server chain + - Output from client chain + - Return to client chain + + + + +[See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9726) + + ### reconnect services dynamically Status: RESOLVED. @@ -94,12 +153,46 @@ Status: RESOLVED. [See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9767) -### IPSec takes a lot of time on request +### questions: dpdk plugin / VCL / service graph topology -Our current implementation uses IKEv2, that requires `rsa` key. -The thing is the key generation takes about 3-5s - https://github.com/networkservicemesh/sdk-vpp/blob/main/pkg/networkservice/mechanisms/ipsec/common.go +Status: RESOLVED. -[See more details](https://github.com/networkservicemesh/sdk-vpp/issues/745) +[See more details](https://github.com/networkservicemesh/deployments-k8s/issues/9777) + + +### Endless retries on error + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/sdk/issues/1500) + + +### Missed red arrows on community page + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/site/issues/256) + + +### Automate updating versing of 'go' + +Updating version of 'go' is a very monotonous and most importantly time-consuming task since we need to update all our repositories (https://github.com/networkservicemesh?q=&type=all&language=go&sort= we have 54 repos) + +[See more details](https://github.com/networkservicemesh/.github/issues/36) + + +### Forwarder sometimes crashes with segmentation fault after restart + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/cmd-forwarder-vpp/issues/953) + + +### netsvcmonitor chain element: nse doesn't match with networkservice + +Status: RESOLVED. + +[See more details](https://github.com/networkservicemesh/sdk/issues/1521)