2918 Commits (bf5b76851a2d6c117e45fe8a9d64c49b010a0c2b)

Author	SHA1	Message	Date
hashicorp-ci	c309d2b6cb	auto-updated agent/uiserver/bindata_assetfs.go from commit 1507dd8ab	4 years ago
hashicorp-ci	61fd873f24	auto-updated agent/uiserver/bindata_assetfs.go from commit 1b5087f6e	4 years ago
Mark Anderson	b9d22f48cd	Add fields to the /acl/auth-methods endpoint. (#9741) ... * A GET of the /acl/auth-method/:name endpoint returns the fields MaxTokenTTL and TokenLocality, while a LIST (/acl/auth-methods) does not. The list command returns a filtered subset of the full set. This is somewhat deliberate, so that secrets aren't shown, but the TTL and Locality fields aren't (IMO) security critical, and it is useful for the front end to be able to show them. For consistency these changes mirror the 'omit empty' and string representation choices made for the GET call. This includes changes to the gRPC and API code in the client. The new output looks similar to this curl 'http://localhost:8500/v1/acl/auth-methods' | jq '.' { "MaxTokenTTL": "8m20s", "Name": "minikube-ttl-local2", "Type": "kubernetes", "Description": "minikube auth method", "TokenLocality": "local", "CreateIndex": 530, "ModifyIndex": 530, "Namespace": "default" } ] Signed-off-by: Mark Anderson <manderson@hashicorp.com> * Add changelog Signed-off-by: Mark Anderson <manderson@hashicorp.com>	4 years ago
Daniel Nephin	d5448c004d	Merge pull request #9452 from hashicorp/dnephin/config-tests-flags-and-edgecases ... config: make TestBuilder_BuildAndValidate_ConfigFlagsAndEdgecases easier to work with	4 years ago
Daniel Nephin	d18e00194a	Merge pull request #9720 from hashicorp/dnephin/ent-meta-ergo-1 ... structs: rename EnterpriseMeta constructor	4 years ago
Daniel Nephin	363d738fd2	Merge pull request #9772 from hashicorp/streamin-fix-bad-cached-snapshot ... streaming: fix snapshot cache bug	4 years ago
Daniel Nephin	89383e2d98	Merge pull request #9728 from hashicorp/dnephin/state-index-table ... state: document how index table is used	4 years ago
Daniel Nephin	d1772ae305	structs: rename EnterpriseMeta constructor ... To match the Go convention.	4 years ago
Daniel Nephin	b91c5a8383	config: rename testCase fields ... So that their name better reflects their purpose	4 years ago
Daniel Nephin	534631659c	config: make TestLoad_IntegrationWithFlags easier to work with ... Replace the large table of tests with individual calls to run(). By using runCase, failure messages will include the line number for the test case, as well as a line number from the test functions. Example: === FAIL: agent/config TestLoad_IntegrationWithFlags/failing_case (0.01s) runtime_test.go:4721: case: failing case runtime_test.go:4864: error "data_dir cannot be empty" does not contain "I expected this error" Previous: runtime_test.go:4864: error "data_dir cannot be empty" does not contain "I expected this error" Without the line number to the testCase data, debugging these tests is difficult. It is impossible to jump directly to the test case, and difficult to find the location because of many similarly named cases.	4 years ago
Daniel Nephin	bd728df345	move test case iteration to caller ... To make the test case logic more obvious	4 years ago
Daniel Nephin	f86ceea803	config: remove test cases for impossible scenarios ... AEInterval is overridden by NonUserSource, so there is no way for a user to set this value. These two cases represented impossible real world scenarios. Instead the test is replaced with one that shows that the AEInterval can not be set by config. This change allows us to remove the hcltail and jsontail fields from testCase	4 years ago
Daniel Nephin	76787d4dd0	Fix error checking ... There were two identical if lines, remove one and use a switch/case so the compiler can catch this problem.	4 years ago
Daniel Nephin	863551beb4	config: move testing shims to LoadOpts ... This will allow us to eventually port these tests to use config.Load instead of unexported functions.	4 years ago
Daniel Nephin	4411da5ddd	Remove patchActual, does not appear to be necessary ... also small rename and fix comment	4 years ago
Daniel Nephin	ba3a1b95e1	stream: fix a snapshot cache bug ... Previously a snapshot created as part of a resumse-stream request could have incorrectly cached the newSnapshotToFollow event. This would cause clients to error because they received an unexpected framing event.	4 years ago
Daniel Nephin	9b3c6da9df	stream: test the snapshot cache is saved correctly ... when the cache entry is created from resuming a stream.	4 years ago
Daniel Nephin	c522f8defd	Merge pull request #9758 from hashicorp/dnephin/fix-streaming-bugs ... http: error if near is used with streaming	4 years ago
Daniel Nephin	0a4304697e	Merge pull request #9727 from mbanikazemi/typos ... Correcting the changed function name in comment	4 years ago
Daniel Nephin	b909088500	http: error if near is used with streaming	4 years ago
hashicorp-ci	bb67e0939a	auto-updated agent/uiserver/bindata_assetfs.go from commit 96204a21d	4 years ago
hashicorp-ci	d5024e64c3	auto-updated agent/uiserver/bindata_assetfs.go from commit 369637aff	4 years ago
hashicorp-ci	673f4690b7	auto-updated agent/uiserver/bindata_assetfs.go from commit eda38a0ba	4 years ago
R.B. Boyer	6eeccc93ce	connect: update supported envoy point releases to 1.16.2, 1.15.3, 1.14.6, 1.13.7 (#9737)	4 years ago
Daniel Nephin	6b36ab744c	streaming: double the cache TTL ... 10 minutes is the default blocking query timeout. Using the same value results in us hitting the expired cache entry bug frequently. By extending this TTL we at least mitigate the problem. The underlying bug still needs to be fixed.	4 years ago
Daniel Nephin	9423c887c4	submatview: do not reset retry waiter when materializer is reset ... The materializer is often reset when an error is received. By resetting the retryWaiter we effectively never wait. The retryWaiter should only be reset when we get an event without error. This is done in Materializer.updateView().	4 years ago
Daniel Nephin	0fa51e5ba9	api: Use blocking query for health when near is set ... Streaming can not be used for these queries because the near query paramter indicates a specific sort of the results, and that sort requires data that is not available to the client from the streaming API.	4 years ago
Matt Keeler	8dbe342ec9	Stop background refresh of cached data for requests that result in ACL not found errors (#9738)	4 years ago
Freddy	82c269a7c5	Avoid potential proxycfg/xDS deadlock using non-blocking send	4 years ago
R.B. Boyer	39e4ae25ac	connect: connect CA Roots in the primary datacenter should use a SigningKeyID derived from their local intermediate (#9428) ... This fixes an issue where leaf certificates issued in primary datacenters using Vault as a Connect CA would be reissued very frequently (every ~20 seconds) because the logic meant to detect root rotation was errantly triggering. The hash of the rootCA was being compared against a hash of the intermediateCA and always failing. This doesn't apply to the Consul built-in CA provider because there is no intermediate in use in the primary DC. This is reminiscent of #6513	4 years ago
freddygv	ec5f75776b	Update comments on avoiding proxycfg deadlock	4 years ago
R.B. Boyer	43193a35c6	xds: prevent LDS flaps in mesh gateways due to unstable datacenter lists (#9651) ... Also fix a similar issue in Terminating Gateways that was masked by an overzealous test.	4 years ago
Mohammad Banikazemi	bcadd341eb	Correcting the changed function name in comment ... Signed-off-by: Mohammad Banikazemi <mbanikazemi@gmail.com>	4 years ago
freddygv	6e443e5536	Retry send after timer fires, in case no updates occur	4 years ago
Daniel Nephin	30332ffb43	state: Use the tableIndex constant	4 years ago
Daniel Nephin	3ecbeda234	state: Document index table ... And move the IndexEntry (which is stored in the table) next to the table schema definition.	4 years ago
R.B. Boyer	adff0c05a7	xds: deduplicate mesh gateway listeners in a stable way (#9650) ... In a situation where the mesh gateway is configured to bind to multiple network interfaces, we use a feature called 'tagged addresses'. Sometimes an address is duplicated across multiple tags such as 'lan' and 'lan_ipv4'. There is code to deduplicate these things when creating envoy listeners, but that code doesn't ensure that the same tag wins every time. If the winning tag flaps between xDS discovery requests it will cause the listener to be drained and replaced.	4 years ago
freddygv	de0cb1af7f	Make xDS labeling consistent with proxycfg	4 years ago
freddygv	95e7641faa	Update proxycfg logging, labels were already attached	4 years ago
Daniel Nephin	a4690ac7d9	Merge pull request #9719 from hashicorp/oss/state-store-4 ... state: remove registerSchema	4 years ago
Daniel Nephin	1c4e0cfa2a	Merge pull request #9718 from hashicorp/oss/dnephin/ent-meta-in-state-store-3 ... state: convert all table name constants to the new prefix pattern	4 years ago
Daniel Nephin	0814f22715	Merge pull request #9665 from hashicorp/dnephin/state-store-indexes-2 ... state: move config-entries table definition to config_entries_schema.go	4 years ago
Daniel Nephin	912dbb4cb4	Merge pull request #9664 from hashicorp/dnephin/state-store-indexes ... state: move ACL schema and index definitions to acl_schema.go	4 years ago
Daniel Nephin	05d5ec4804	state: remove the need for registerSchema ... registerSchema creates some indirection which is not necessary in this case. newDBSchema can call each of the tables. Enterprise tables can be added from the existing withEnterpriseSchema shim.	4 years ago
Daniel Nephin	2cbf8b5fd0	state: rename table name constants to use pattern ... the 'table' prefix is shorter, and also reads better in queries.	4 years ago
Daniel Nephin	8ac9d54ccc	state: rename connect constants	4 years ago
Daniel Nephin	0c34e474c5	state: rename table name constants to new pattern ... Using Apps Hungarian Notation for these constants makes the memdb queries more readable.	4 years ago
Daniel Nephin	2d5b5afec1	state: Remove unnecessary entMeta arg to EnsureConfigEntry	4 years ago
freddygv	5ba14ad41d	Add trace logs to proxycfg state runner and xds srv	4 years ago
freddygv	37190c0d0d	Avoid potential deadlock using non-blocking send ... Deadlock scenario: 1. Due to scheduling, the state runner sends one snapshot into snapCh and then attempts to send a second. The first send succeeds because the channel is buffered, but the second blocks. 2. Separately, Manager.Watch is called by the xDS server after getting a discovery request from Envoy. This function acquires the manager lock and then blocks on receiving the CurrentSnapshot from the state runner. 3. Separately, there is a Manager goroutine that reads the snapshots from the channel in step 1. These reads are done to notify proxy watchers, but they require holding the manager lock. This goroutine goes to acquire that lock, but can't because it is held by step 2. Now, the goroutine from step 3 is waiting on the one from step 2 to release the lock. The goroutine from step 2 won't release the lock until the goroutine in step 1 advances. But the goroutine in step 1 is waiting for the one in step 3. Deadlock. By making this send non-blocking step 1 above can proceed. The coalesce timer will be reset and a new valid snapshot will be delivered after it elapses or when one is requested by xDS.	4 years ago