cluster/kube/k0.libsonnet

// k0.hswaw.net kubernetes cluster
// This defines the cluster as a single object.
// Use the sibling k0*.jsonnet 'view' files to actually apply the configuration.

local kube = import "../../kube/kube.libsonnet";
local policies = import "../../kube/policies.libsonnet";

local cluster = import "cluster.libsonnet";

local admitomatic = import "lib/admitomatic.libsonnet";
local cockroachdb = import "lib/cockroachdb.libsonnet";
local registry = import "lib/registry.libsonnet";
local rook = import "lib/rook.libsonnet";
local admins = import "lib/admins.libsonnet";

{
    k0: {
        local k0 = self,
        cluster: cluster.Cluster("k0", "hswaw.net") {
            cfg+: {
                storageClassNameParanoid: k0.ceph.waw3Pools.blockRedundant.name,
            },
            metallb+: {
                cfg+: {
                    // Peer with calico running on same node.
                    peers: [
                        {
                            "peer-address": "127.0.0.1",
                            "peer-asn": 65003,
                            "my-asn": 65002,
                        },
                    ],
                    // Public IP address pools. Keep in sync with k0.calico.yaml.
                    addressPools: [
                        {
                            name: "public-v4-1",
                            protocol: "bgp",
                            addresses: [
                                "185.236.240.48/28",
                            ],
                        },
                        {
                            name: "public-v4-2",
                            protocol: "bgp",
                            addresses: [
                                "185.236.240.112/28"
                            ],
                        },
                    ],
                },
            },
        },

        // Docker registry
        registry: registry.Environment {
            cfg+: {
                domain: "registry.%s" % [k0.cluster.fqdn],
                storageClassName: k0.cluster.cfg.storageClassNameParanoid,
                objectStorageName: "waw-hdd-redundant-3-object",
            },
        },

        // CockroachDB, running on bc01n{01,02,03}.
        cockroach: {
            waw2: cockroachdb.Cluster("crdb-waw1") {
                cfg+: {
                    topology: [
                        { name: "dcr01s22", node: "dcr01s22.hswaw.net" },
                        { name: "dcr01s24", node: "dcr01s24.hswaw.net" },
                        { name: "dcr03s16", node: "dcr03s16.hswaw.net" },
                    ],
                    // Host path on SSD.
                    hostPath: "/var/db/crdb-waw1",
                    extraDNS: [
                        "crdb-waw1.hswaw.net",
                    ],
                },
                initJob:: null,
            },
            clients: {
                cccampix: k0.cockroach.waw2.Client("cccampix"),
                cccampixDev: k0.cockroach.waw2.Client("cccampix-dev"),
                buglessDev: k0.cockroach.waw2.Client("bugless-dev"),
                sso: k0.cockroach.waw2.Client("sso"),
                herpDev: k0.cockroach.waw2.Client("herp-dev"),
                gitea: k0.cockroach.waw2.Client("gitea"),
                issues: k0.cockroach.waw2.Client("issues"),
                dns: k0.cockroach.waw2.Client("dns"),
            },
        },

        ceph: {
            // waw1 cluster - dead as of 2019/08/06, data corruption
            // waw2 cluster - dead as of 2021/01/22, torn down (horrible M610 RAID controllers are horrible)

            // waw3: 6TB SAS 3.5" HDDs, internal Rook cluster.
            //
            // Suffers from rook going apeshit and nuking all mons if enough of
            // a control plane is up for rook to run but if nodes are
            // unavailable to the point of it deciding that no mon exists and
            // it should create some new ones, fully nuking the monmap and
            // making recovery a pain.
            //
            // Supposedly new versions of Rook slowly fix these issues, but q3k
            // doesn't personally trust this codebase anymore. He'd rather
            // manage the actual Ceph cluster myself, we don't need all of this
            // magic.
            //
            // See: b.hswaw.net/6
            waw3: rook.Cluster(k0.cluster.rook, "ceph-waw3") {
                spec: {
                    mon: {
                        count: 3,
                        allowMultiplePerNode: false,
                    },
                    resources: {
                        osd: {
                            requests: {
                                cpu: "2",
                                memory: "6G",
                            },
                            limits: {
                                cpu: "2",
                                memory: "8G",
                            },
                        },

                    },
                    storage: {
                        useAllNodes: false,
                        useAllDevices: false,
                        config: {
                            databaseSizeMB: "1024",
                            journalSizeMB: "1024",
                        },

                        nodes: [
                            {
                                name: "dcr01s22.hswaw.net",
                                devices: [
                                    { name: "/dev/disk/by-id/wwn-0x" + id }
                                    for id in [
                                        "5000c5008508c433", # ST6000NM0034 Z4D40QZR0000R629ME1B
                                        "5000c500850989cf", # ST6000NM0034 Z4D40JRL0000R63008A2
                                        "5000c5008508baf7", # ST6000NM0034 Z4D40M380000R630V00M
                                        "5000c5008508f843", # ST6000NM0034 Z4D40LGP0000R630UVTD
                                        "5000c500850312cb", # ST6000NM0034 Z4D3ZAAX0000R629NW31
                                        "5000c500850293e3", # ST6000NM0034 Z4D3Z5TD0000R629MF7P
                                        "5000c5008508e3ef", # ST6000NM0034 Z4D40LM50000R630V0W3
                                        "5000c5008508e23f", # ST6000NM0034 Z4D40QMX0000R629MD3C
                                    ]
                                ],
                            },
                            {
                                name: "dcr01s24.hswaw.net",
                                devices: [
                                    { name: "/dev/disk/by-id/wwn-0x" + id }
                                    for id in [
                                        "5000c5008508c9ef", # ST6000NM0034 Z4D40LY40000R630UZCE
                                        "5000c5008508df33", # ST6000NM0034 Z4D40QQ00000R629MB25
                                        "5000c5008508dd3b", # ST6000NM0034 Z4D40QQJ0000R630RBY6
                                        "5000c5008509199b", # ST6000NM0034 Z4D40QG10000R630V0X9
                                        "5000c5008508ee03", # ST6000NM0034 Z4D40LHH0000R630UYP0
                                        "5000c50085046abf", # ST6000NM0034 Z4D3ZF1B0000R629NV9P
                                        "5000c5008502929b", # ST6000NM0034 Z4D3Z5WG0000R629MF14
                                    ]
                                ],
                            },
                            {
                                name: "dcr03s16.hswaw.net",
                                devices: [
                                    { name: "/dev/disk/by-id/wwn-0x" + id }
                                    for id in [
                                        "5000c5008508fb73", # ST6000NM0034 Z4D40LEF0000R630UX98
                                        "5000c5008508c3a7", # ST6000NM0034 Z4D40LZV0000R630UY91
                                        "5000c5008508d7bf", # ST6000NM0034 Z4D40LPT0000R629NXBF
                                        "5000c5008502952f", # ST6000NM0034 Z4D3Z5RA0000R628P45F
                                        "5000c5008502aa4b", # ST6000NM0034 Z4D3Z5A00000R630RU2T
                                        "5000c5008508d677", # ST6000NM0034 Z4D40LQH0000R630QRAS
                                    ]
                                ],
                            },
                        ],
                    },
                    benji:: {
                        metadataStorageClass: "waw-hdd-redundant-3",
                        encryptionPassword: std.split((importstr "../secrets/plain/k0-benji-encryption-password"), '\n')[0],
                        pools: [
                            "waw-hdd-redundant-3",
                            "waw-hdd-redundant-3-metadata",
                            "waw-hdd-yolo-3",
                        ],
                        s3Configuration: {
                            awsAccessKeyId: "RPYZIROFXNLQVU2WJ4R3",
                            awsSecretAccessKey: std.split((importstr "../secrets/plain/k0-benji-secret-access-key"), '\n')[0],
                            bucketName: "benji-k0-backups-waw3",
                            endpointUrl: "https://s3.eu-central-1.wasabisys.com/",
                        },
                    }
                },
            },
            waw3Pools: {
                // redundant block storage
                blockRedundant: rook.ECBlockPool(k0.ceph.waw3, "waw-hdd-redundant-3") {
                    metadataReplicas: 2,
                    spec: {
                        failureDomain: "host",
                        replicated: {
                          size: 2,
                        },
                    },
                },
                // q3k's personal pool, used externally from k8s.
                q3kRedundant: rook.ECBlockPool(k0.ceph.waw3, "waw-hdd-redundant-q3k-3") {
                    metadataReplicas: 2,
                    spec: {
                        failureDomain: "host",
                        replicated: {
                          size: 2,
                        },
                    },
                },

                object: {
                    local poolSpec = {
                        failureDomain: "host",
                        replicated: { size: 2 },
                    },

                    realm: rook.S3ObjectRealm(k0.ceph.waw3, "hscloud"),
                    zonegroup: rook.S3ObjectZoneGroup(self.realm, "eu"),
                    // This is serving at object.ceph-waw3.hswaw.net, but
                    // internally to Ceph it is known as
                    // waw-hdd-redundant-3-object (name of radosgw zone).
                    store: rook.S3ObjectStore(self.zonegroup, "waw-hdd-redundant-3-object") {
                        cfg+: {
                            // Override so that this radosgw serves on
                            // object.ceph-{waw3,eu}.hswaw.net instead of
                            // ceph-{waw-hdd-redundant-3-object,eu}.
                            domainParts: [
                                "waw3", "eu",
                            ],
                        },
                        spec: {
                            metadataPool: poolSpec,
                            dataPool: poolSpec,
                        },
                    },
                },
            },

            // Clients for S3/radosgw storage.
            clients: {
                local ObjectStoreUser(name) = kube.CephObjectStoreUser(name) {
                    metadata+: {
                        namespace: "ceph-waw3",
                    },
                    spec: {
                        store: "waw-hdd-redundant-3-object",
                        displayName: name,
                    },
                },

                # Used for owncloud.hackerspace.pl, which for now lives on boston-packets.hackerspace.pl.
                nextcloudWaw3: ObjectStoreUser("nextcloud"),
                # issues.hackerspace.pl (redmine) attachments bucket
                issuesWaw3: ObjectStoreUser("issues"),
                # matrix.hackerspace.pl media storage bucket
                matrixWaw3: ObjectStoreUser("matrix"),
                # tape staging temporary storage
                tapeStaging: ObjectStoreUser("tape-staging"),

                # nuke@hackerspace.pl's personal storage.
                nukePersonalWaw3: ObjectStoreUser("nuke-personal"),

                # patryk@hackerspace.pl's ArmA3 mod bucket.
                cz2ArmaModsWaw3: ObjectStoreUser("cz2-arma3mods"),

                # implr's personal user
                implrSparkWaw3: ObjectStoreUser("implr"),

                # q3k's personal user
                q3kWaw3: ObjectStoreUser("q3k"),

                # woju's personal user
                wojuWaw3: ObjectStoreUser("woju"),

                # cz3's (patryk@hackerspace.pl) personal user
                cz3Waw3: ObjectStoreUser("cz3"),

                # informatic's personal user
                informaticWaw3: ObjectStoreUser("informatic"),

                # mastodon qa and prod
                mastodonWaw3: {
                    qa: ObjectStoreUser("mastodon-qa"),
                    prod: ObjectStoreUser("mastodon-prod"),
                },

                codehostingWaw3: ObjectStoreUser("codehosting"),
            },
        },


        # These are policies allowing for Insecure pods in some namespaces.
        # A lot of them are spurious and come from the fact that we deployed
        # these namespaces before we deployed the draconian PodSecurityPolicy
        # we have now. This should be fixed by setting up some more granular
        # policies, or fixing the workloads to not need some of the permission
        # bits they use, whatever those might be.
        # TODO(q3k): fix this?
        unnecessarilyInsecureNamespaces: [
            policies.AllowNamespaceInsecure("ceph-waw3"),
            policies.AllowNamespaceInsecure("matrix"),
            policies.AllowNamespaceInsecure("registry"),
            policies.AllowNamespaceInsecure("internet"),
            # TODO(implr): restricted policy with CAP_NET_ADMIN and tuntap, but no full root
            policies.AllowNamespaceInsecure("implr-vpn"),
            // For SourceGraph's tini container mess.
            policies.AllowNamespaceMostlySecure("sourcegraph"),
        ],

        # Admission controller that permits non-privileged users to manage
        # their namespaces without danger of hijacking important URLs.
        admitomatic: admitomatic.Environment {
            cfg+: {
                proto: {
                    // Domains allowed in given namespaces. If a domain exists
                    // anywhere, ingresses will only be permitted to be created
                    // within namespaces in which it appears here. This works
                    // the same way for wildcards, if a wildcard exists in this
                    // list it blocks all unauthorized uses of that domain
                    // elsewhere.
                    //
                    // See //cluster/admitomatic for more information.
                    //
                    // Or, tl;dr:
                    //
                    // If you do a wildcard CNAME onto the k0 ingress, you
                    // should explicitly state *.your.name.com here.
                    //
                    // If you just want to protect your host from being
                    // hijacked by other cluster users, you should also state
                    // it here (either as a wildcard, or unary domains).
                    allow_domain: [
                        { namespace: "inventory", dns: "inventory.hackerspace.pl" },
                        { namespace: "capacifier", dns: "capacifier.hackerspace.pl" },
                        { namespace: "ldapweb", dns: "profile.hackerspace.pl" },
                        { namespace: "walne", dns: "walne.hackerspace.pl" },
                        { namespace: "devtools-prod", dns: "hackdoc.hackerspace.pl" },
                        { namespace: "devtools-prod", dns: "cs.hackerspace.pl" },
                        { namespace: "sourcegraph", dns: "cs.hackerspace.pl" },
                        { namespace: "codehosting-prod", dns: "git.hackerspace.pl" },
                        { namespace: "codehosting-prod", dns: "code.hackerspace.pl" },
                        { namespace: "engelsystem-prod", dns: "engelsystem.hackerspace.pl" },
                        { namespace: "gerrit", dns: "gerrit.hackerspace.pl" },
                        { namespace: "gerrit-qa", dns: "gerrit-qa.hackerspace.pl" },
                        { namespace: "gitea-prod", dns: "gitea.hackerspace.pl" },
                        { namespace: "hswaw-prod", dns: "*.hackerspace.pl" },
                        { namespace: "hswaw-prod", dns: "*.hswaw.net" },
                        { namespace: "site", dns: "new.hackerspace.pl" },
                        { namespace: "internet", dns: "internet.hackerspace.pl" },
                        { namespace: "matrix", dns: "matrix.hackerspace.pl" },
                        { namespace: "onlyoffice-prod", dns: "office.hackerspace.pl" },
                        { namespace: "paperless", dns: "paperless.hackerspace.pl" },
                        { namespace: "redmine", dns: "issues.hackerspace.pl" },
                        { namespace: "redmine", dns: "b.hackerspace.pl" },
                        { namespace: "redmine", dns: "b.hswaw.net" },
                        { namespace: "redmine", dns: "xn--137h.hackerspace.pl" },
                        { namespace: "redmine", dns: "xn--137h.hswaw.net" },
                        { namespace: "speedtest", dns: "speedtest.hackerspace.pl" },
                        { namespace: "sso", dns: "sso.hackerspace.pl" },
                        { namespace: "mastodon-hackerspace-qa", dns: "social-qa-2.hackerspace.pl" },
                        { namespace: "mastodon-hackerspace-prod", dns: "social.hackerspace.pl" },

                        // auto-namespaced domains, i.e:
                        // USER.hscloud.ovh is allowed for personal-USER namespace
                        // *.USER.hscloud.ovh is allowed for personal-USER namespace
                        { namespace: "personal-$2", dns: "(.*\\.)?([^.]+)\\.hscloud\\.ovh", regexp: true },

                        // cluster infra
                        { namespace: "ceph-waw3", dns: "ceph-waw3.hswaw.net" },
                        { namespace: "ceph-waw3", dns: "object.ceph-waw3.hswaw.net" },
                        { namespace: "ceph-waw3", dns: "object.ceph-eu.hswaw.net" },
                        { namespace: "monitoring-global-k0", dns: "*.hswaw.net" },
                        { namespace: "registry", dns: "*.hswaw.net" },

                        // personal namespaces
                        { namespace: "q3k", dns: "*.q3k.org" },
                        { namespace: "personal-q3k", dns: "*.q3k.org" },
                        { namespace: "personal-radex", dns: "hs.radex.io" },
                        { namespace: "personal-radex", dns: "*.hs.radex.io" },
                    ],

                    anything_goes_namespace: [
                        // sourcegraph ingress wants a config snippet to set a header.
                        "devtools-prod",
                        "sourcegraph",
                    ],
                },
            },
        },

        // Configuration of RoleBindings
        admins: admins.NamespaceAdmins {
            // Cluster staff have full access to all namespaces
            // To give non-staff users admin access scoped to a given namespace,
            // add them to the list below.
            // (system:admin-namespace role is given to <user>@hackerspace.pl)
            namespaces:: {
                "inventory": [
                    "radex",
                    "palid",
                ],
                "walne": [
                    "radex",
                    "palid",
                ],
                "site": [
                    "ar",
                    "radex",
                ],
                "valheim": [
                    "patryk",
                    "palid",
                ],
                "matrix-0x3c": [
                    "not7cd",
                ],
                "hswaw-prod": [
                    "ar",
                    "radex",
                ],
                "ldapweb": [
                    "radex",
                ],
                "devtools-prod": [
                    "radex",
                ],
                "depotview": [
                    "radex",
                ],
                "hackdoc": [
                    "radex",
                ],
                "sourcegraph": [
                    "radex",
                ],
                "speedtest": [
                    "radex",
                ],
                "internet": [
                    "radex",
                ],
                "cebulacamp": [
                    "radex",
                ],
                "teleimg": [
                    "radex",
                ],
                "pretalx": [
                    "radex",
                ],
            }
        }
    },
}