Fix deploy Kubernetes cluster with workers #3870

0oM4R · 2025-02-06T16:34:32Z

The issue that checking if the node is definded and already created breaks the creation of the workers, it passes with master as there is no network was created yet so it will continue normally then to the workers will face the break line before adding the workers

Changes

k8s: for workers and masters only check if the network is loaded if not create a new network without break;

Related Issues

🐞 [Bug]: Can't deploy a k8s cluster with workers [Only master got deployed] #3863

Tested Scenarios

deployed a cluster with 2 workers from dashboard

deployed cluster using kubernetes_leader script
passed already created network and the cluster deployed normally on it

Documentation PR

For UI changes, Please provide the Documentation PR on info_grid

To consider

Preliminary Checks:

Preliminary Checks
- Does it completely address the issue linked?
- What about edge cases?
- Does it meet the specified acceptance criteria?
- Are there any unintended side effects?
- Does the PR adhere to the team's coding conventions, style guides, and best practices?
- Does it integrate well with existing features?
- Does it impact the overall performance of the application?
- Are there any bottlenecks or slowdowns?
- Has it been optimized for efficiency?
- Has it been adequately tested with unit, integration, and end-to-end tests?
- Are there any known defects or issues?
- Is the code well-documented?
- Are changes to documentation reflected in the code?

UI Checks:

UI Checks
- If a UI design is provided/ does it follow it?
- Does every button work?
- Is the data displayed logical? Is it what you expected?
- Does every validation work?
- Does this interface feel intuitive?
- What about slow network? Offline?
- What if the gridproxy/graphql/chain is failing?
- What would a first time user have a hard time navigating here?

Code Quality Checks:

Code Quality Checks
- Code formatted/linted? Are there unused imports? .. etc
- Is there redundant/repeated code?
- Are there conditionals that are always true or always false?
- Can we write this more concisely?
- Can we reuse this code? If yes, where?
- Will the changes be easy to maintain and update in the future?
- Can this code become too complex to understand for other devs?
- Can this code cause future integration problems?

Testing Checklist

Does it Meet the specified acceptance criteria.
Test if changes can affect any other functionality.
Tested with unit, integration, and end-to-end tests.
Tested the un-happy path and rollback scenarios.
Documentation updated to meet the latest changes.
Check automated tests working successfully with the changes.
Can be covered by automated tests.

General Checklist

- k8s: for workers and masters only check if the network is loaded if not create a new network without break;

amiraabouhadid

works, good job
{ "masters": [ { "version": 0, "type": "zmachine", "contractId": 191373, "nodeId": 153, "name": "mrtyklo", "created": 1739174873, "status": "ok", "message": "", "flist": "https://hub.grid.tf/tf-official-apps/threefolddev-k3s-v1.31.0.flist", "publicIP": null, "myceliumIP": "42e:525:4c9e:f4e4:ff0f:0:400:2", "interfaces": [ { "network": "nwkies6", "ip": "10.20.4.2" } ], "capacity": { "cpu": 1, "memory": 4096 }, "mounts": [ { "name": "mrtyklo_disk", "mountPoint": "/mnt/data", "size": 107374182400, "state": "ok", "message": "" } ], "env": { "SSH_KEY": "ssh-rsa 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 cosmos", "K3S_TOKEN": "YiEwYq98B6", "K3S_DATA_DIR": "/mnt/data", "K3S_FLANNEL_IFACE": "eth0", "K3S_NODE_NAME": "mrtyklo", "K3S_URL": "" }, "entrypoint": "/sbin/zinit init", "metadata": "", "description": "", "rootfs_size": 524288000, "corex": false, "gpu": [] } ], "workers": [ { "version": 0, "type": "zmachine", "contractId": 191372, "nodeId": 128, "name": "wrcjlrr", "created": 1739174874, "status": "ok", "message": "", "flist": "https://hub.grid.tf/tf-official-apps/threefolddev-k3s-v1.31.0.flist", "publicIP": null, "myceliumIP": "554:d273:2e79:9ec6:ff0f:0:500:2", "interfaces": [ { "network": "nwkies6", "ip": "10.20.5.2" } ], "capacity": { "cpu": 1, "memory": 4096 }, "mounts": [ { "name": "wrcjlrr_disk", "mountPoint": "/mnt/data", "size": 107374182400, "state": "ok", "message": "" } ], "env": { "SSH_KEY": "ssh-rsa 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 cosmos", "K3S_TOKEN": "YiEwYq98B6", "K3S_DATA_DIR": "/mnt/data", "K3S_FLANNEL_IFACE": "eth0", "K3S_NODE_NAME": "wrcjlrr", "K3S_URL": "https://10.20.4.2:6443" }, "entrypoint": "/sbin/zinit init", "metadata": "", "description": "", "rootfs_size": 524288000, "corex": false, "gpu": [] } ], "wireguard": "[Interface]\nAddress = 100.64.20.3/32\nPrivateKey = x/ugVx7GragYRHRZE/lQ+dZRBXx6RE21ge74J/c35hM=\n\n[Peer]\nPublicKey = XugpkuEvSbnBwDkHyP+y3juqzCjmuQ23EmaWkRy+uTk=\nAllowedIPs = 10.20.0.0/16, 100.64.20.0/32\nPersistentKeepalive = 25\nEndpoint = 185.206.122.31:17621" }

AhmedHanafy725 · 2025-02-10T08:20:09Z

packages/grid_client/src/modules/k8s.ts

      // Sets the network and contractMetadata based on the node's zos version
      const nodeTwinId = await this.capacity.getNodeTwinId(master.node_id);
      const features = await this.rmb.request([nodeTwinId], "zos.system.node_features_get", "", 20, 3);
      if (features.some(item => item.includes("zmachine-light") || item.includes("network-light"))) {
-        network = new ZNetworkLight(options.network.name, options.network.ip_range, this.config);
-        await network.load();
+        if (!network) {


Will this network be defined at any point?

please check that in the rest of the changes

Will this network be defined at any point?

Yes it will be defined in the next iteration also in the workers

AhmedHanafy725 · 2025-02-10T08:24:34Z

packages/grid_client/src/modules/k8s.ts

      if (features.some(item => item.includes("zmachine-light") || item.includes("network-light"))) {
-        network = new ZNetworkLight(options.network.name, options.network.ip_range, this.config);
-        await network.load();
+        if (!network) {
+          network = new ZNetworkLight(options.network.name, options.network.ip_range, this.config);
+          await network.load();
+        }
        contractMetadata = JSON.stringify({
          version: 4,
          type: "kubernetes",
          name: options.name,
          projectName: this.config.projectName || `kubernetes/${options.name}`,
        });
      } else {
-        network = new Network(options.network.name, options.network.ip_range, this.config);
-        await network.load();
+        if (!network) {
+          network = new Network(options.network.name, options.network.ip_range, this.config);
+          await network.load();
+        }


why is this part repeated?

I tried to create a function in utils that does those steps, but the function will take the network argument, network, and network type, so there will be more lines

example

awat createOrLoadNetowrk(options.network.name, options.network.ip_range, this.config, network, 'ZNetworkLight`)

the case you might have only workers if you add a worker to an existing cluster

fix(K8s[workers]):

74fc4fb

- k8s: for workers and masters only check if the network is loaded if not create a new network without break;

0oM4R requested review from AhmedHanafy725, zaelgohary, maayarosama, MohamedElmdary, Mahmoud-Emad, amiraabouhadid, AlaaElattar and samaradel as code owners February 6, 2025 16:34

maayarosama approved these changes Feb 9, 2025

View reviewed changes

amiraabouhadid approved these changes Feb 10, 2025

View reviewed changes

AhmedHanafy725 reviewed Feb 10, 2025

View reviewed changes

AhmedHanafy725 approved these changes Feb 10, 2025

View reviewed changes

0oM4R merged commit 1c2d461 into development Feb 10, 2025
9 checks passed

0oM4R deleted the development_fix_deploy_k8sWorkers branch February 10, 2025 09:12

0oM4R mentioned this pull request Feb 10, 2025

🐞 [Bug]: Received contracts not as expected #3847

Closed

1 task

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix deploy Kubernetes cluster with workers #3870

Fix deploy Kubernetes cluster with workers #3870

0oM4R commented Feb 6, 2025 •

edited

Loading

amiraabouhadid left a comment

AhmedHanafy725 Feb 10, 2025

AhmedHanafy725 Feb 10, 2025

0oM4R Feb 10, 2025

AhmedHanafy725 Feb 10, 2025

0oM4R Feb 10, 2025 •

edited

Loading

AhmedHanafy725 Feb 10, 2025

Fix deploy Kubernetes cluster with workers #3870

Fix deploy Kubernetes cluster with workers #3870

Conversation

0oM4R commented Feb 6, 2025 • edited Loading

Changes

Related Issues

Tested Scenarios

Documentation PR

To consider

Preliminary Checks:

UI Checks:

Code Quality Checks:

Testing Checklist

General Checklist

amiraabouhadid left a comment

Choose a reason for hiding this comment

AhmedHanafy725 Feb 10, 2025

Choose a reason for hiding this comment

AhmedHanafy725 Feb 10, 2025

Choose a reason for hiding this comment

0oM4R Feb 10, 2025

Choose a reason for hiding this comment

AhmedHanafy725 Feb 10, 2025

Choose a reason for hiding this comment

0oM4R Feb 10, 2025 • edited Loading

Choose a reason for hiding this comment

AhmedHanafy725 Feb 10, 2025

Choose a reason for hiding this comment

0oM4R commented Feb 6, 2025 •

edited

Loading

0oM4R Feb 10, 2025 •

edited

Loading