cli, ui: dismiss release notes signup banner per environment variable #56437

nkodali · 2020-11-09T15:34:45Z

Previously, the signup banner could only be dismissed manually.
For internal testing purposes, this banner is unnecessary. This
change provides a way to dismiss the signup banner upon start of
a cluster via the cli by setting the environment variable
COCKROACH_UI_RELEASE_NOTES_SIGNUP_DISMISSED=true.

Resolves #46998

Release note: none

cockroach-teamcity · 2020-11-09T15:34:53Z

This change is

dhartunian

Reviewed 3 of 3 files at r1.
Reviewable status: complete! 1 of 0 LGTMs obtained

knz

Reviewable status: complete! 0 of 0 LGTMs obtained (and 1 stale) (waiting on @dhartunian and @nkodali)

pkg/cli/start.go, line 607 at r2 (raw file):

			// Configure UI settings.
			cc, _, finish, err := getClientGRPCConn(ctx, serverCfg)

No that is not a valid approach technically. We do not guarantee the admin server to be available for an internal loopback connection in this way.

getClientGRPCConn is meant for use in a client command running in a different process, possibly over the network. Some server configurations will entirely prevent a client connection from being established from the same host.

Here the proper approach is to take the code in pkg/server/admin.go from the SetUIData() method handler, and move it to a differetn (new) function, then export that function, then use that function directly from here.

nkodali

Reviewable status: complete! 0 of 0 LGTMs obtained (and 1 stale) (waiting on @dhartunian and @knz)

pkg/cli/start.go, line 607 at r2 (raw file):

Previously, knz (kena) wrote…

No that is not a valid approach technically. We do not guarantee the admin server to be available for an internal loopback connection in this way.

getClientGRPCConn is meant for use in a client command running in a different process, possibly over the network. Some server configurations will entirely prevent a client connection from being established from the same host.

Here the proper approach is to take the code in pkg/server/admin.go from the SetUIData() method handler, and move it to a differetn (new) function, then export that function, then use that function directly from here.

Done. I did notice some of the other CLI commands (e.g. init, debug, node, quit) get client connections to issue various requests, sometimes even to admin server. Is there something that makes those situations different than this one? Trying to understand when it is safe to get a client connection vs doing what you suggest here and interfacing directly with the server side code. Thanks for the help!

knz

Thanks this looks better.

Reviewed 1 of 3 files at r1, 3 of 3 files at r3.
Reviewable status: complete! 0 of 0 LGTMs obtained (and 1 stale) (waiting on @nkodali)

pkg/cli/start.go, line 607 at r2 (raw file):

Previously, nkodali wrote…

Done. I did notice some of the other CLI commands (e.g. init, debug, node, quit) get client connections to issue various requests, sometimes even to admin server. Is there something that makes those situations different than this one? Trying to understand when it is safe to get a client connection vs doing what you suggest here and interfacing directly with the server side code. Thanks for the help!

These are client commands that run in a different process, and meant to be run over the network. The code here is part of the server process.

pkg/server/admin.go, line 1201 at r3 (raw file):

// This function writes to system.ui using the internal SQL executor of the provided server.
func SetUIData(
	ctx context.Context, s *Server, req *serverpb.SetUIDataRequest,

It's unclear to me why you pass Server as argument here. Why not make this a method on *Server like the other one? You can give it a different name.

pkg/server/admin.go, line 1203 at r3 (raw file):

	ctx context.Context, s *Server, req *serverpb.SetUIDataRequest,
) (*serverpb.SetUIDataResponse, error) {
	ctx = s.AnnotateCtx(ctx)

please move this ctx assignment to the SetUIData() method above - it belongs to RPC requests.

pkg/server/admin.go, line 1204 at r3 (raw file):

) (*serverpb.SetUIDataResponse, error) {
	ctx = s.AnnotateCtx(ctx)
	userName, err := userFromContext(ctx)

Same, the userName belongs to the RPC server. Pass the username as argument to this function/method.

nkodali

Reviewable status: complete! 0 of 0 LGTMs obtained (and 1 stale) (waiting on @knz)

pkg/server/admin.go, line 1201 at r3 (raw file):

Previously, knz (kena) wrote…

It's unclear to me why you pass Server as argument here. Why not make this a method on *Server like the other one? You can give it a different name.

Done.

pkg/server/admin.go, line 1203 at r3 (raw file):

Previously, knz (kena) wrote…

please move this ctx assignment to the SetUIData() method above - it belongs to RPC requests.

Done.

pkg/server/admin.go, line 1204 at r3 (raw file):

Previously, knz (kena) wrote…

Same, the userName belongs to the RPC server. Pass the username as argument to this function/method.

Done.

knz

thanks!

Reviewed 2 of 2 files at r4.
Reviewable status: complete! 1 of 0 LGTMs obtained (and 1 stale) (waiting on @knz)

nkodali · 2020-11-12T03:38:09Z

I'm having some trouble getting all the roachtests to pass. I ran the roachtests from my machine and everything succeeds. I've re-run on TeamCity several times now. The first few times, different tests failed each time, leading to me believe there is some flakiness perhaps? In later runs, acceptance/bank/cluster-recovery test was failing for a few runs in a row. Any suggestions to debug further?

knz · 2020-11-12T13:49:42Z

yes that looks like flakiness that's not yours. can you try rebasing once more maybe?

knz · 2020-11-12T15:01:01Z

that rebase went wrong. there's now too much in the pr

nkodali · 2020-11-12T19:59:04Z

bors r=knz,dhartunian

craig · 2020-11-12T21:07:49Z

Build failed (retrying...):

GitHub CI (Cockroach)

Previously, when starting a node using roachprod, the db console ui would display a release notes signup banner. This banner is not useful for users of roachprod. We introduced a way to toggle this via env var in cockroachdb#56437. This commit sets that env var, COCKROACH_UI_RELEASE_NOTES_SIGNUP_DISMISSED=true for `cockroach start*` commands issued from roachprod. Resolves: cockroachdb#46998 See also: cockroachdb#56437 Release note: none

craig · 2020-11-13T01:44:05Z

Build failed (retrying...):

GitHub CI (Cockroach)

craig · 2020-11-13T03:46:49Z

Build failed (retrying...):

GitHub CI (Cockroach)

@sumeerbhola

56373: hlc: introduce synthetic flag on timestamps r=nvanbenschoten a=nvanbenschoten Informs #52745. Informs #36431. This commit introduces an 8-bit `flags` field on the hlc timestamp struct. The flags are used to provide details about the timestamp and its meaning. They do not affect the sort order of Timestamps. The commit then introduces the first flag: SYNTHETIC. As discussed in #52745, a synthetic timestamp is defined as a timestamp that makes no claim about the value of clocks in the system. While standard timestamps are pulled from HLC clocks and indicate that some node in the system has a clock with a reading equal to or above its value, a synthetic timestamp makes no such indication. By avoiding a connection to "real time", synthetic timestamps can be used to write values at a future time and to indicate that observed timestamps do not apply to such writes for the purposes of tracking causality between the write and its observers. Observed timestamps will be a critical part of implementing non-blocking transactions (#52745) and fixing the interaction between observed timestamps and transaction refreshing (#36431). The original plan was to reserve the high-order bit in the logical portion of a timestamp as a "synthetic bit". This is how I began implementing things, but was turned off for a few reasons. First, it was fairly subtle and seemed too easy to get wrong. Using a separate field is more explicit and avoids a class of bugs. Second, I began to have serious concerns about how the synthetic bit would impact timestamp ordering. Every timestamp comparison would need to mask out the bit or risk being incorrect. This was even true of the LSM custom comparator. This seemed difficult to get right and seemed particularly concerning since we're planning on marking only some of a transaction's committed values as synthetic to fix #36431, so if we weren't careful, we could get atomicity violations. There were also minor backwards compatibility concerns. But a separate field is more expensive in theory, so we need to be careful. However, it turns out that a separate field is mostly free in each case that we care about. In memory, the separate field is effectively free because the Timestamp struct was previously 12 bytes but was always padded out to 16 bytes when included as a field in any other struct. This means that the flags field is replacing existing padding. Over the wire, the field will not be included when zero and will use a varint encoding when not zero, so again, it is mostly free. In the engine key encoding, the field is also not included when zero, and takes up only 1 byte when non-zero, so it is mostly free. ---- First three commits from #56477. @sumeerbhola I'm hoping you can take a look at the engine-level changes in the `introduce synthetic flag on timestamps` commit (4th commit as of the time of writing). I think the key encoding added here makes sense, but want to make sure you're on board. One possible concern is that we introduce a new 13-byte suffix, which means that combined with a 4-byte sequence number (see #41720 (comment)), we'd collide with the 17 byte `engineKeyVersionLockTableLen`. @tbg do you mind being the primary reviewer here? I think you know the most about the motivations for this change and will have a good sense of whether this is the best way to introduce additional state on timestamps. 56437: cli, ui: dismiss release notes signup banner per environment variable r=knz,dhartunian a=nkodali Previously, the signup banner could only be dismissed manually. For internal testing purposes, this banner is unnecessary. This change provides a way to dismiss the signup banner upon start of a cluster via the cli by setting the environment variable COCKROACH_UI_RELEASE_NOTES_SIGNUP_DISMISSED=true. Resolves #46998 Release note: none 56627: sql: rework SHOW REGIONS to SHOW REGIONS FROM CLUSTER r=ajstorm a=otan Resolves #56331 Release note (sql change): SHOW REGIONS functionality is now deferred to SHOW REGIONS FROM CLUSTER. Co-authored-by: Nathan VanBenschoten <[email protected]> Co-authored-by: Namrata Kodali <[email protected]> Co-authored-by: Oliver Tan <[email protected]>

craig · 2020-11-13T05:51:27Z

Build failed (retrying...):

GitHub CI (Cockroach)

craig · 2020-11-13T07:40:21Z

Build failed (retrying...):

GitHub CI (Cockroach)

craig · 2020-11-13T08:43:24Z

Build failed:

GitHub CI (Cockroach)

knz · 2020-11-13T10:01:46Z

@nkodali I believe we have identified the source of the test flake so this may not be able to merge until we fix that.

@tbg here's an example PR that's blocked on the node restart. I believe that's what you aim to fix with the error refactor?

Previously, when starting a node using roachprod, the db console ui would display a release notes signup banner. This banner is not useful for users of roachprod. We introduced a way to toggle this via env var in cockroachdb#56437. This commit sets that env var, COCKROACH_UI_RELEASE_NOTES_SIGNUP_DISMISSED=true for `cockroach start*` commands issued from roachprod. Resolves: cockroachdb#46998 See also: cockroachdb#56437 Release note: none

nkodali

@knz thanks for the heads up

Reviewable status: complete! 0 of 0 LGTMs obtained (and 2 stale) (waiting on @dhartunian and @knz)

knz · 2020-11-18T11:14:51Z

could you rebase this today and see what CI thinks about it now? Thanks

Previously, the signup banner could only be dismissed manually. For internal testing purposes, this banner is unnecessary. This change provides a way to dismiss the signup banner upon start of a cluster via the cli by setting the environment variable COCKROACH_UI_RELEASE_NOTES_SIGNUP_DISMISSED=true. Resolves cockroachdb#46998 Release note: none

nkodali · 2020-11-18T18:09:50Z

bors r=knz,dhartunian

craig · 2020-11-18T18:51:45Z

Build failed:

GitHub CI (Cockroach)

knz · 2020-11-18T18:58:43Z

ok either you are being unlucky, or there is a real bug here underneath.

@bdarnell can we solicit your additional review here? Is there something in this PR that could cause a lot of our acceptance tests to fail with RPC or HTTP errorS?

nkodali · 2020-11-20T16:52:27Z

bors r=knz,dhartunian

craig · 2020-11-20T17:51:01Z

Build succeeded:

GitHub CI (Cockroach)

56632: roachprod: dismiss signup banner for nodes started by roachprod r=dhartunian,irfansharif a=nkodali Previously, when starting a node using roachprod, the db console ui would display a release notes signup banner. This banner is not useful for users of roachprod. We introduced a way to toggle this via env var in #56437. This commit sets that env var, COCKROACH_UI_RELEASE_NOTES_SIGNUP_DISMISSED=true for `cockroach start*` commands issued from roachprod. Resolves: #46998 See also: #56437 Release note: none 56913: sqlproxyccl: count successful connection attempts r=spaskob a=spaskob This counter will be useful to understand how many successful connections have been cached so far unless we have gone over the cache size limit of course. Release note: none. Co-authored-by: Namrata Kodali <[email protected]> Co-authored-by: Spas Bojanov <[email protected]>

nkodali requested a review from dhartunian November 9, 2020 15:34

nkodali requested a review from a team as a code owner November 9, 2020 15:34

dhartunian approved these changes Nov 9, 2020

View reviewed changes

nkodali force-pushed the signup-banner-env-var branch from 4eb0e7f to 66db602 Compare November 9, 2020 17:24

knz suggested changes Nov 10, 2020

View reviewed changes

nkodali force-pushed the signup-banner-env-var branch from 66db602 to f69d6bb Compare November 11, 2020 01:28

nkodali commented Nov 11, 2020

View reviewed changes

knz reviewed Nov 11, 2020

View reviewed changes

nkodali force-pushed the signup-banner-env-var branch from f69d6bb to 32e1802 Compare November 11, 2020 15:00

nkodali commented Nov 11, 2020

View reviewed changes

knz approved these changes Nov 11, 2020

View reviewed changes

nkodali requested review from a team November 12, 2020 14:41

nkodali requested a review from a team as a code owner November 12, 2020 14:41

nkodali requested review from dt and removed request for a team November 12, 2020 14:41

nkodali marked this pull request as draft November 12, 2020 14:45

dhartunian removed the request for review from a team November 12, 2020 15:06

nkodali force-pushed the signup-banner-env-var branch from 3283eab to 55327e5 Compare November 12, 2020 15:18

nkodali marked this pull request as ready for review November 12, 2020 15:19

nkodali removed the request for review from dt November 12, 2020 15:25

nkodali force-pushed the signup-banner-env-var branch from 55327e5 to 7604831 Compare November 12, 2020 16:48

nkodali commented Nov 13, 2020

View reviewed changes

nkodali force-pushed the signup-banner-env-var branch from 7604831 to dd55f54 Compare November 17, 2020 18:55

nkodali force-pushed the signup-banner-env-var branch from dd55f54 to dd9613c Compare November 18, 2020 16:44

craig bot merged commit 2d88439 into cockroachdb:master Nov 20, 2020

jordanlewis mentioned this pull request Nov 22, 2020

bank: cluster-recovery and node-restart are throughly flaky or broken #57002

Closed

nkodali deleted the signup-banner-env-var branch November 24, 2020 22:19

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

cli, ui: dismiss release notes signup banner per environment variable #56437

cli, ui: dismiss release notes signup banner per environment variable #56437

nkodali commented Nov 9, 2020

cockroach-teamcity commented Nov 9, 2020

dhartunian left a comment

knz left a comment

nkodali left a comment

knz left a comment

nkodali left a comment

knz left a comment

nkodali commented Nov 12, 2020

knz commented Nov 12, 2020

knz commented Nov 12, 2020

nkodali commented Nov 12, 2020

craig bot commented Nov 12, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

knz commented Nov 13, 2020

nkodali left a comment

knz commented Nov 18, 2020

nkodali commented Nov 18, 2020

craig bot commented Nov 18, 2020

knz commented Nov 18, 2020

nkodali commented Nov 20, 2020

craig bot commented Nov 20, 2020

cli, ui: dismiss release notes signup banner per environment variable #56437

cli, ui: dismiss release notes signup banner per environment variable #56437

Conversation

nkodali commented Nov 9, 2020

cockroach-teamcity commented Nov 9, 2020

dhartunian left a comment

Choose a reason for hiding this comment

knz left a comment

Choose a reason for hiding this comment

nkodali left a comment

Choose a reason for hiding this comment

knz left a comment

Choose a reason for hiding this comment

nkodali left a comment

Choose a reason for hiding this comment

knz left a comment

Choose a reason for hiding this comment

nkodali commented Nov 12, 2020

knz commented Nov 12, 2020

knz commented Nov 12, 2020

nkodali commented Nov 12, 2020

craig bot commented Nov 12, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

craig bot commented Nov 13, 2020

knz commented Nov 13, 2020

nkodali left a comment

Choose a reason for hiding this comment

knz commented Nov 18, 2020

nkodali commented Nov 18, 2020

craig bot commented Nov 18, 2020

knz commented Nov 18, 2020

nkodali commented Nov 20, 2020

craig bot commented Nov 20, 2020