You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
consul/agent/structs/acl_test.go

704 lines
17 KiB

package structs
import (
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
"fmt"
"strings"
"testing"
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
"github.com/hashicorp/consul/acl"
"github.com/stretchr/testify/require"
)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
func TestStructs_ACLToken_PolicyIDs(t *testing.T) {
t.Parallel()
t.Run("Basic", func(t *testing.T) {
t.Parallel()
token := &ACLToken{
Policies: []ACLTokenPolicyLink{
ACLTokenPolicyLink{
ID: "one",
},
ACLTokenPolicyLink{
ID: "two",
},
ACLTokenPolicyLink{
ID: "three",
},
},
}
policyIDs := token.PolicyIDs()
require.Len(t, policyIDs, 3)
require.Equal(t, "one", policyIDs[0])
require.Equal(t, "two", policyIDs[1])
require.Equal(t, "three", policyIDs[2])
})
t.Run("Legacy Management", func(t *testing.T) {
t.Parallel()
a := &ACL{
ID: "root",
Type: ACLTokenTypeManagement,
Name: "management",
}
token := a.Convert()
policyIDs := token.PolicyIDs()
require.Len(t, policyIDs, 0)
embedded := token.EmbeddedPolicy()
require.NotNil(t, embedded)
require.Equal(t, ACLPolicyGlobalManagement, embedded.Rules)
})
t.Run("Legacy Management With Rules", func(t *testing.T) {
t.Parallel()
a := &ACL{
ID: "root",
Type: ACLTokenTypeManagement,
Name: "management",
Rules: "operator = \"write\"",
}
token := a.Convert()
policyIDs := token.PolicyIDs()
require.Len(t, policyIDs, 0)
embedded := token.EmbeddedPolicy()
require.NotNil(t, embedded)
require.Equal(t, ACLPolicyGlobalManagement, embedded.Rules)
})
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
t.Run("No Policies", func(t *testing.T) {
t.Parallel()
token := &ACLToken{}
policyIDs := token.PolicyIDs()
require.Len(t, policyIDs, 0)
})
}
func TestStructs_ACLToken_EmbeddedPolicy(t *testing.T) {
t.Parallel()
t.Run("No Rules", func(t *testing.T) {
t.Parallel()
token := &ACLToken{}
require.Nil(t, token.EmbeddedPolicy())
})
t.Run("Legacy Client", func(t *testing.T) {
t.Parallel()
// None of the other fields should be considered
token := &ACLToken{
Type: ACLTokenTypeClient,
Rules: `acl = "read"`,
}
policy := token.EmbeddedPolicy()
require.NotNil(t, policy)
require.NotEqual(t, "", policy.ID)
require.True(t, strings.HasPrefix(policy.Name, "legacy-policy-"))
require.Equal(t, token.Rules, policy.Rules)
require.Equal(t, policy.Syntax, acl.SyntaxLegacy)
require.NotNil(t, policy.Hash)
require.NotEqual(t, []byte{}, policy.Hash)
})
t.Run("Same Policy for Tokens with same Rules", func(t *testing.T) {
t.Parallel()
token1 := &ACLToken{
AccessorID: "f55b260c-5e05-418e-ab19-d421d1ab4b52",
SecretID: "b2165bac-7006-459b-8a72-7f549f0f06d6",
Description: "token 1",
Type: ACLTokenTypeClient,
Rules: `acl = "read"`,
}
token2 := &ACLToken{
AccessorID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
SecretID: "65e98e67-9b29-470c-8ffa-7c5a23cc67c8",
Description: "token 2",
Type: ACLTokenTypeClient,
Rules: `acl = "read"`,
}
policy1 := token1.EmbeddedPolicy()
policy2 := token2.EmbeddedPolicy()
require.Equal(t, policy1, policy2)
})
}
func TestStructs_ACLServiceIdentity_SyntheticPolicy(t *testing.T) {
t.Parallel()
for _, test := range []struct {
serviceName string
datacenters []string
expectRules string
}{
{"web", nil, `
service "web" {
policy = "write"
}
service "web-sidecar-proxy" {
policy = "write"
}
service_prefix "" {
policy = "read"
}
node_prefix "" {
policy = "read"
}`},
{"companion-cube-99", []string{"dc1", "dc2"}, `
service "companion-cube-99" {
policy = "write"
}
service "companion-cube-99-sidecar-proxy" {
policy = "write"
}
service_prefix "" {
policy = "read"
}
node_prefix "" {
policy = "read"
}`},
} {
name := test.serviceName
if len(test.datacenters) > 0 {
name += " [" + strings.Join(test.datacenters, ", ") + "]"
}
t.Run(name, func(t *testing.T) {
svcid := &ACLServiceIdentity{
ServiceName: test.serviceName,
Datacenters: test.datacenters,
}
expect := &ACLPolicy{
Syntax: acl.SyntaxCurrent,
Datacenters: test.datacenters,
Description: "synthetic policy",
Rules: test.expectRules,
}
got := svcid.SyntheticPolicy()
require.NotEmpty(t, got.ID)
require.True(t, strings.HasPrefix(got.Name, "synthetic-policy-"))
// strip irrelevant fields before equality
got.ID = ""
got.Name = ""
got.Hash = nil
require.Equal(t, expect, got)
})
}
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
func TestStructs_ACLToken_SetHash(t *testing.T) {
t.Parallel()
token := ACLToken{
AccessorID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
SecretID: "65e98e67-9b29-470c-8ffa-7c5a23cc67c8",
Description: "test",
Policies: []ACLTokenPolicyLink{
ACLTokenPolicyLink{
ID: "one",
},
ACLTokenPolicyLink{
ID: "two",
},
ACLTokenPolicyLink{
ID: "three",
},
},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
t.Run("Nil Hash - Generate", func(t *testing.T) {
require.Nil(t, token.Hash)
h := token.SetHash(false)
require.NotNil(t, h)
require.NotEqual(t, []byte{}, h)
require.Equal(t, h, token.Hash)
})
t.Run("Hash Set - Dont Generate", func(t *testing.T) {
original := token.Hash
h := token.SetHash(false)
require.Equal(t, original, h)
token.Description = "changed"
h = token.SetHash(false)
require.Equal(t, original, h)
})
t.Run("Hash Set - Generate", func(t *testing.T) {
original := token.Hash
h := token.SetHash(true)
require.NotEqual(t, original, h)
})
}
func TestStructs_ACLToken_EstimateSize(t *testing.T) {
t.Parallel()
// estimated size here should
token := ACLToken{
AccessorID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
SecretID: "65e98e67-9b29-470c-8ffa-7c5a23cc67c8",
Description: "test",
Policies: []ACLTokenPolicyLink{
ACLTokenPolicyLink{
ID: "one",
},
ACLTokenPolicyLink{
ID: "two",
},
ACLTokenPolicyLink{
ID: "three",
},
},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
// this test is very contrived. Basically just tests that the
// math is okay and returns the value.
require.Equal(t, 128, token.EstimateSize())
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
}
func TestStructs_ACLToken_Stub(t *testing.T) {
t.Parallel()
t.Run("Basic", func(t *testing.T) {
t.Parallel()
token := ACLToken{
AccessorID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
SecretID: "65e98e67-9b29-470c-8ffa-7c5a23cc67c8",
Description: "test",
Policies: []ACLTokenPolicyLink{
ACLTokenPolicyLink{
ID: "one",
},
ACLTokenPolicyLink{
ID: "two",
},
ACLTokenPolicyLink{
ID: "three",
},
},
}
stub := token.Stub()
require.Equal(t, token.AccessorID, stub.AccessorID)
require.Equal(t, token.Description, stub.Description)
require.Equal(t, token.Policies, stub.Policies)
require.Equal(t, token.Local, stub.Local)
require.Equal(t, token.CreateTime, stub.CreateTime)
require.Equal(t, token.Hash, stub.Hash)
require.Equal(t, token.CreateIndex, stub.CreateIndex)
require.Equal(t, token.ModifyIndex, stub.ModifyIndex)
require.False(t, stub.Legacy)
})
t.Run("Legacy", func(t *testing.T) {
t.Parallel()
token := ACLToken{
AccessorID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
SecretID: "65e98e67-9b29-470c-8ffa-7c5a23cc67c8",
Description: "test",
Type: ACLTokenTypeClient,
Rules: `key "" { policy = "read" }`,
}
stub := token.Stub()
require.Equal(t, token.AccessorID, stub.AccessorID)
require.Equal(t, token.Description, stub.Description)
require.Equal(t, token.Policies, stub.Policies)
require.Equal(t, token.Local, stub.Local)
require.Equal(t, token.CreateTime, stub.CreateTime)
require.Equal(t, token.Hash, stub.Hash)
require.Equal(t, token.CreateIndex, stub.CreateIndex)
require.Equal(t, token.ModifyIndex, stub.ModifyIndex)
require.True(t, stub.Legacy)
})
}
func TestStructs_ACLTokens_Sort(t *testing.T) {
t.Parallel()
tokens := ACLTokens{
&ACLToken{
AccessorID: "9db509a9-c809-48c1-895d-99f845b7a9d5",
},
&ACLToken{
AccessorID: "6bd01084-1695-43b8-898d-b2dd7874754d",
},
&ACLToken{
AccessorID: "614a4cef-9149-4271-b878-7edb1ad661f8",
},
&ACLToken{
AccessorID: "c9dd9980-8d54-472f-9e5e-74c02143e1f4",
},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
tokens.Sort()
require.Equal(t, tokens[0].AccessorID, "614a4cef-9149-4271-b878-7edb1ad661f8")
require.Equal(t, tokens[1].AccessorID, "6bd01084-1695-43b8-898d-b2dd7874754d")
require.Equal(t, tokens[2].AccessorID, "9db509a9-c809-48c1-895d-99f845b7a9d5")
require.Equal(t, tokens[3].AccessorID, "c9dd9980-8d54-472f-9e5e-74c02143e1f4")
}
func TestStructs_ACLTokenListStubs_Sort(t *testing.T) {
t.Parallel()
tokens := ACLTokenListStubs{
&ACLTokenListStub{
AccessorID: "9db509a9-c809-48c1-895d-99f845b7a9d5",
},
&ACLTokenListStub{
AccessorID: "6bd01084-1695-43b8-898d-b2dd7874754d",
},
&ACLTokenListStub{
AccessorID: "614a4cef-9149-4271-b878-7edb1ad661f8",
},
&ACLTokenListStub{
AccessorID: "c9dd9980-8d54-472f-9e5e-74c02143e1f4",
},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
tokens.Sort()
require.Equal(t, tokens[0].AccessorID, "614a4cef-9149-4271-b878-7edb1ad661f8")
require.Equal(t, tokens[1].AccessorID, "6bd01084-1695-43b8-898d-b2dd7874754d")
require.Equal(t, tokens[2].AccessorID, "9db509a9-c809-48c1-895d-99f845b7a9d5")
require.Equal(t, tokens[3].AccessorID, "c9dd9980-8d54-472f-9e5e-74c02143e1f4")
}
func TestStructs_ACLPolicy_Stub(t *testing.T) {
t.Parallel()
policy := &ACLPolicy{
ID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
Name: "test",
Description: "test",
Rules: `acl = "read"`,
}
stub := policy.Stub()
require.Equal(t, policy.ID, stub.ID)
require.Equal(t, policy.Name, stub.Name)
require.Equal(t, policy.Description, stub.Description)
require.Equal(t, policy.Datacenters, stub.Datacenters)
require.Equal(t, policy.Hash, stub.Hash)
require.Equal(t, policy.CreateIndex, stub.CreateIndex)
require.Equal(t, policy.ModifyIndex, stub.ModifyIndex)
}
func TestStructs_ACLPolicy_SetHash(t *testing.T) {
t.Parallel()
policy := &ACLPolicy{
ID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
Name: "test",
Description: "test",
Rules: `acl = "read"`,
}
t.Run("Nil Hash - Generate", func(t *testing.T) {
require.Nil(t, policy.Hash)
h := policy.SetHash(false)
require.NotNil(t, h)
require.NotEqual(t, []byte{}, h)
require.Equal(t, h, policy.Hash)
})
t.Run("Hash Set - Dont Generate", func(t *testing.T) {
original := policy.Hash
h := policy.SetHash(false)
require.Equal(t, original, h)
policy.Description = "changed"
h = policy.SetHash(false)
require.Equal(t, original, h)
})
t.Run("Hash Set - Generate", func(t *testing.T) {
original := policy.Hash
h := policy.SetHash(true)
require.NotEqual(t, original, h)
})
}
func TestStructs_ACLPolicy_EstimateSize(t *testing.T) {
t.Parallel()
policy := ACLPolicy{
ID: "09d1c059-961a-46bd-a2e4-76adebe35fa5",
Name: "test",
Description: "test",
Rules: `acl = "read"`,
}
// this test is very contrived. Basically just tests that the
// math is okay and returns the value.
require.Equal(t, 84, policy.EstimateSize())
policy.Datacenters = []string{"dc1", "dc2"}
require.Equal(t, 90, policy.EstimateSize())
}
func TestStructs_ACLPolicies_Sort(t *testing.T) {
t.Parallel()
policies := ACLPolicies{
&ACLPolicy{
ID: "9db509a9-c809-48c1-895d-99f845b7a9d5",
},
&ACLPolicy{
ID: "6bd01084-1695-43b8-898d-b2dd7874754d",
},
&ACLPolicy{
ID: "614a4cef-9149-4271-b878-7edb1ad661f8",
},
&ACLPolicy{
ID: "c9dd9980-8d54-472f-9e5e-74c02143e1f4",
},
}
policies.Sort()
require.Equal(t, policies[0].ID, "614a4cef-9149-4271-b878-7edb1ad661f8")
require.Equal(t, policies[1].ID, "6bd01084-1695-43b8-898d-b2dd7874754d")
require.Equal(t, policies[2].ID, "9db509a9-c809-48c1-895d-99f845b7a9d5")
require.Equal(t, policies[3].ID, "c9dd9980-8d54-472f-9e5e-74c02143e1f4")
}
func TestStructs_ACLPolicyListStubs_Sort(t *testing.T) {
t.Parallel()
policies := ACLPolicyListStubs{
&ACLPolicyListStub{
ID: "9db509a9-c809-48c1-895d-99f845b7a9d5",
},
&ACLPolicyListStub{
ID: "6bd01084-1695-43b8-898d-b2dd7874754d",
},
&ACLPolicyListStub{
ID: "614a4cef-9149-4271-b878-7edb1ad661f8",
},
&ACLPolicyListStub{
ID: "c9dd9980-8d54-472f-9e5e-74c02143e1f4",
},
}
policies.Sort()
require.Equal(t, policies[0].ID, "614a4cef-9149-4271-b878-7edb1ad661f8")
require.Equal(t, policies[1].ID, "6bd01084-1695-43b8-898d-b2dd7874754d")
require.Equal(t, policies[2].ID, "9db509a9-c809-48c1-895d-99f845b7a9d5")
require.Equal(t, policies[3].ID, "c9dd9980-8d54-472f-9e5e-74c02143e1f4")
}
func TestStructs_ACLPolicies_resolveWithCache(t *testing.T) {
t.Parallel()
config := ACLCachesConfig{
Identities: 0,
Policies: 0,
ParsedPolicies: 4,
Authorizers: 0,
Roles: 0,
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
}
cache, err := NewACLCaches(&config)
require.NoError(t, err)
testPolicies := ACLPolicies{
&ACLPolicy{
ID: "5d5653a1-2c2b-4b36-b083-fc9f1398eb7b",
Name: "policy1",
Description: "policy1",
Rules: `node_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 1,
ModifyIndex: 2,
},
},
&ACLPolicy{
ID: "b35541f0-a88a-48da-bc66-43553c60b628",
Name: "policy2",
Description: "policy2",
Rules: `agent_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 3,
ModifyIndex: 4,
},
},
&ACLPolicy{
ID: "383abb79-94ca-46c6-89b7-8ecb69046de9",
Name: "policy3",
Description: "policy3",
Rules: `key_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 5,
ModifyIndex: 6,
},
},
&ACLPolicy{
ID: "8bf38965-95e5-4e86-9be7-f6070cc0708b",
Name: "policy4",
Description: "policy4",
Rules: `service_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 7,
ModifyIndex: 8,
},
},
}
t.Run("Cache Misses", func(t *testing.T) {
policies, err := testPolicies.resolveWithCache(cache, nil)
require.NoError(t, err)
require.Len(t, policies, 4)
for i := range testPolicies {
require.Equal(t, testPolicies[i].ID, policies[i].ID)
require.Equal(t, testPolicies[i].ModifyIndex, policies[i].Revision)
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
})
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
t.Run("Check Cache", func(t *testing.T) {
for i := range testPolicies {
entry := cache.GetParsedPolicy(fmt.Sprintf("%x", testPolicies[i].Hash))
require.NotNil(t, entry)
require.Equal(t, testPolicies[i].ID, entry.Policy.ID)
require.Equal(t, testPolicies[i].ModifyIndex, entry.Policy.Revision)
// set this to detect using from the cache next time
entry.Policy.Revision = 9999
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
})
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
t.Run("Cache Hits", func(t *testing.T) {
policies, err := testPolicies.resolveWithCache(cache, nil)
require.NoError(t, err)
require.Len(t, policies, 4)
for i := range testPolicies {
require.Equal(t, testPolicies[i].ID, policies[i].ID)
require.Equal(t, uint64(9999), policies[i].Revision)
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
})
}
func TestStructs_ACLPolicies_Compile(t *testing.T) {
t.Parallel()
config := ACLCachesConfig{
Identities: 0,
Policies: 0,
ParsedPolicies: 4,
Authorizers: 2,
Roles: 0,
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
}
cache, err := NewACLCaches(&config)
require.NoError(t, err)
testPolicies := ACLPolicies{
&ACLPolicy{
ID: "5d5653a1-2c2b-4b36-b083-fc9f1398eb7b",
Name: "policy1",
Description: "policy1",
Rules: `node_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 1,
ModifyIndex: 2,
},
},
&ACLPolicy{
ID: "b35541f0-a88a-48da-bc66-43553c60b628",
Name: "policy2",
Description: "policy2",
Rules: `agent_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 3,
ModifyIndex: 4,
},
},
&ACLPolicy{
ID: "383abb79-94ca-46c6-89b7-8ecb69046de9",
Name: "policy3",
Description: "policy3",
Rules: `key_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 5,
ModifyIndex: 6,
},
},
&ACLPolicy{
ID: "8bf38965-95e5-4e86-9be7-f6070cc0708b",
Name: "policy4",
Description: "policy4",
Rules: `service_prefix "" { policy = "read" }`,
Syntax: acl.SyntaxCurrent,
RaftIndex: RaftIndex{
CreateIndex: 7,
ModifyIndex: 8,
},
},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
6 years ago
t.Run("Cache Miss", func(t *testing.T) {
authz, err := testPolicies.Compile(acl.DenyAll(), cache, nil)
require.NoError(t, err)
require.NotNil(t, authz)
require.True(t, authz.NodeRead("foo"))
require.True(t, authz.AgentRead("foo"))
require.True(t, authz.KeyRead("foo"))
require.True(t, authz.ServiceRead("foo"))
require.False(t, authz.ACLRead())
})
t.Run("Check Cache", func(t *testing.T) {
entry := cache.GetAuthorizer(testPolicies.HashKey())
require.NotNil(t, entry)
authz := entry.Authorizer
require.NotNil(t, authz)
require.True(t, authz.NodeRead("foo"))
require.True(t, authz.AgentRead("foo"))
require.True(t, authz.KeyRead("foo"))
require.True(t, authz.ServiceRead("foo"))
require.False(t, authz.ACLRead())
// setup the cache for the next test
cache.PutAuthorizer(testPolicies.HashKey(), acl.DenyAll())
})
t.Run("Cache Hit", func(t *testing.T) {
authz, err := testPolicies.Compile(acl.DenyAll(), cache, nil)
require.NoError(t, err)
require.NotNil(t, authz)
// we reset the Authorizer in the cache so now everything should be denied
require.False(t, authz.NodeRead("foo"))
require.False(t, authz.AgentRead("foo"))
require.False(t, authz.KeyRead("foo"))
require.False(t, authz.ServiceRead("foo"))
require.False(t, authz.ACLRead())
})
}