* Randomize order of CT logs when submitting precerts so we maximize the chances we actually exercise all of the logs in a group and not just the first in the list.
* Add metrics for winning logs
When the WFE calls the RA the RA creates a sub context which is cancelled when
the RPC returns. Because we were spawning the publisher RPC calls in a
goroutine with the context from ra.IssueCertificate as soon as
ra.IssueCertificate returned that context was being canceled which in turn
canceled the publisher RPC calls.
Instead of using the RA RPC context simply use a `context.Background()` so
that the RPC context doesn't break these submissions. Also return to
pre-features.CancelCTSubmissions behavior where precert submissions would
be canceled once we retrieved SCTs from the winning logs instead of relying
on the magic behavior of the RA RPC canceling them itself.
The Boulder orphan-finder command uses the SA's AddCertificate RPC to add orphaned certificates it finds back to the DB. Prior to this commit this RPC always set the core.Certificate.Issued field to the
current time. For the orphan-finder case this meant that the Issued date would incorrectly be set to when the certificate was found, not when it was actually issued. This could cause cert-checker to alarm based on the unusual delta between the cert NotBefore and the core.Certificate.Issued value.
This PR updates the AddCertificate RPC to accept an optional issued timestamp in the request arguments. In the SA layer we address deployability concerns by setting a default value of the current time when none is explicitly provided. This matches the classic behaviour and will let an old RA communicate with a new SA.
This PR updates the orphan-finder to provide an explicit issued time to sa.AddCertificate. The explicit issued time is calculated using the found certificate's NotBefore and the configured backdate.
This lets the orphan-finder set the true issued time in the core.Certificate object, avoiding any cert-checker alarms.
Resolves#3624
This PR updates the Boulder github.com/weppos/publicsuffix-go dependency to
weppos/publicsuffix-go@542377b - the tip of master at the time of writing.
Unit tests are confirmed to pass:
$ go test ./...
? github.com/weppos/publicsuffix-go/cmd/load [no test files]
ok github.com/weppos/publicsuffix-go/net/publicsuffix 0.005s
ok github.com/weppos/publicsuffix-go/publicsuffix 0.022s
Notably this update adds the .sport TLD and we've had some requests to support issuance for domains under this newly created TLD.
This allows us to have fast-running unittests without modifying the global state in singleDialTimeout,
which can become a const.
Fixes#3628.
Builds on top of #3629, review that first.
In particular, differentiate timeouts during connect (which are usually a firewall problem) from timeouts after connect (which are usually a software problem). In the process, refactor the tests and add testing for specific problem detail messages.
This also switches over the HTTP challenge's dialer to use DialContext, and to shave a little bit of headroom off of the context deadline, so that the dial can report its timeout before the overall context expires, which would lead to an overly generic "deadline exceeded" error, which would then get translated (incorrectly) into a "timeout after connect."
There is an additional error case, Timeout during %s (your server may be slow or overloaded), (where %s can be read or write) which doesn't have any unittests. I believe it may not be possible to trigger this, since read and write timeouts get subsumed by the HTTP or TLS library, but it's worth having as a fallback case. We'll see if it shows up in the logs.
Among the test refactorings, I shortened the timeout on the TLS timeout test to 50ms. Previously this was the long pole making the whole test take 10s. Now it takes ~500 ms overall.
I recommend starting review at https://github.com/letsencrypt/boulder/compare/detailed-va-errors?expand=1#diff-4c51d1d7ca3ec3022d14b42809af0d7eR671 (the changes to detailedError), then reviewing the Dial -> DialContext changes, then the tests.
Submits final certificates to any configured CT logs. This doesn't introduce a feature flag as it is config gated, any log we want to submit final certificates to needs to have it's log description updated to include the `"submitFinalCerts": true` field.
Fixes#3605.
During periods of peak load, some RPCs are significantly delayed (on the order of seconds) by client-side blocking. HTTP/2 clients have to obey a "max concurrent streams" setting sent by the server. In Go's HTTP/2 implementation, this value [defaults to 250](https://github.com/golang/net/blob/master/http2/server.go#L56), so the gRPC default is also 250. So whenever there are more than 250 requests in progress at a time, additional requests will be delayed until there is a slot available.
During this peak load, we aren't hitting limits on CPU or memory, so we should increase the max concurrent streams limit to take better advantage of our available resources. This PR adds a config field to do that.
Fixes#3641.
In #3614 we adjusted the `SA.NewOrder` function to conditionally call `ssa.statusForOrder` on the new order when `features.OrderReadyStatus` was enabled. Unfortunately this call to `ssa.statusForOrder` happened *before* the `req.BeganProcessing` field was initialized with a pointer to a `false` bool. The `ssa.statusForOrder` function (correctly) assumes that `req.BeganProcessing == nil` is illegal and doesn't correspond to a known status. This results in `NewOrder` requests returning a 500 error
of the form:
> Internal error - Error creating new order - Order XXX is in an invalid state. No state known for this order's authorizations
Our integration tests missed this because we didn't have a test case that issued for a set of names with one account, and then issued again for the same set of names with the same account.
This PR fixes the original bug by moving the `BeganProcessing` initialization before the call to `statusForOrder`. This PR also adds an integration test to catch this sort of bug again in the future.
Prior to the SA fix this test failed with the 500 server internal error observed by the Certbot team. With the SA fix in place the test passes again.
Finally, this PR disables the `OrderReadyStatus` feature flag in `test/config-next/sa.json`. Certbot's ACME implementation breaks when this flag is enabled (See https://github.com/certbot/certbot/issues/5856). Since Certbot runs integration tests against Boulder with config-next we should be courteous and leave this flag disabled until we are closer to being able to turn it on for staging/prod.
The `TotalCertificates` rate limit serves to ensure we don't
accidentally exceed our OCSP signing capacity by issuing too many
certificates within a fixed period. In practice this rate limit has been
fragile and the associated queries have been linked to performance
problems.
Since we now have better means of monitoring our OCSP signing capacity
this commit removes the rate limit and associated code.
We have updated staging/prod Boulder builds to use Go 1.10.1. This means
we no longer need to support Go 1.10.0 in dev docker images, CI, and our
image building tools.
This commit disables the `OrderReadyStatus` feature flag in
`test/config-next/sa.json`. Certbot's ACME implementation breaks when
this flag is enabled (See
https://github.com/certbot/certbot/issues/5856). Since Certbot runs
integration tests against Boulder with config-next we should be
courteous and leave this flag disabled until we are closer to being able
to turn it on for staging/prod.
In #3614 we adjusted the `SA.NewOrder` function to conditionally call
`ssa.statusForOrder` on the new order when `features.OrderReadyStatus`
was enabled. Unfortunately this call to `ssa.statusForOrder` happened
*before* the `req.BeganProcessing` field was initialized with a pointer
to a `false` bool. The `ssa.statusForOrder` function (correctly) assumes
that `req.BeganProcessing == nil` is illegal and doesn't correspond to
a known status. This results in NewOrder requests returning a 500 error
of the form:
> Internal error - Error creating new order - Order XXX is in an invalid
> state. No state known for this order's authorizations
Our integration tests missed this because we didn't have a test case
that issued for a set of names with one account, and then issued again
for the same set of names with the same account.
This commit fixes the original bug by moving the `BeganProcessing`
initialization before the call to `statusForOrder`. This commit also
adds an integration test to catch this sort of bug again in the
future.
Prior to the SA fix this test failed with the 500 server internal error
observed by the Certbot team. With the SA fix in place the test passes
again.
* SA: Add Order "Ready" status, feature flag.
This commit adds the new "Ready" status to `core/objects.go` and updates
`sa.statusForOrder` to use it conditionally for orders with all valid
authorizations that haven't been finalized yet. This state is used
conditionally based on the `features.OrderReadyStatus` feature flag
since it will likely break some existing clients that expect status
"Processing" for this state. The SA unit test for `statusForOrder` is
updated with a "ready" status test case.
* RA: Enforce order ready status conditionally.
This commit updates the RA to conditionally expect orders that are being
finalized to be in the "ready" status instead of "pending". This is
conditionally enforced based on the `OrderReadyStatus` feature flag.
Along the way the SA was changed to calculate the order status for the
order returned in `sa.NewOrder` dynamically now that it could be
something other than "pending".
* WFE2: Conditionally enforce order ready status for finalization.
Similar to the RA the WFE2 should conditionally enforce that an order's
status is either "ready" or "pending" based on the "OrderReadyStatus"
feature flag.
* Integration: Fix `test_order_finalize_early`.
This commit updates the V2 `test_order_finalize_early` test for the
"ready" status. A nice side-effect of the ready state change is that we
no longer invalidate an order when it is finalized too soon because we
can reject the finalization in the WFE. Subsequently the
`test_order_finalize_early` testcase is also smaller.
* Integration: Test classic behaviour w/o feature flag.
In the previous commit I fixed the integration test for the
`config/test-next` run that has the `OrderReadyStatus` feature flag set
but broke it for the `config/test` run without the feature flag.
This commit updates the `test_order_finalize_early` test to work
correctly based on the feature flag status in both cases.
Previously we updated the RA's issueCertificateInner function to prefix errors returned from the CA with meaningful information about which CA RPC caused the failure. Unfortunately by using fmt.Errorf to do this we're discarding the underlying error type. This can cause unexpected server internal errors downstream if (for e.g.) the CA rejects a CSR with a malformed error (see #3632).
This PR updates the issueCertificateInner error message prefixing to maintain the error type if the underlying error is a berrors.BoulderError. A RA unit test with several mock CAs is added to test the prefixing occurs as expected without loss of error type.
This PR also adds an integration test that ensures we reject a CSR with >100 names with a malformed error. This is not strictly related to this PR but since I wrote it while debugging the root issue I thought I'd include it. To allow this test to pass the pendingAuthorizationsPerAccount in test/rate-limit-policies.yml and associated tests had to be adjusted.
Resolves#3632
This commit updates the `boulder-ra` and `boulder-ca` commands to refuse
to start if their configured `MaxNames` is 0 (the default value). This
should always be set to a positive number.
This commit also updates `csr/csr.go` to always apply the max names
check since it will never be 0 after the change above.
Also refactor `FailOnError` to pull out a separate `Fail` function.
Related to https://github.com/letsencrypt/boulder/issues/3632
Distinguish between deadline exceeded vs canceled. Also, combine those
two cases with "out of retries" into a single stat with a label
determining type.
gRPC passes deadline information through the RPC boundary, but client and server have the same deadline. Ideally we'd like the server to have a slightly tighter deadline than the client, so if one of the server's onward RPCs or other network calls times out, the server can pass back more detailed information to the client, rather than the client timing out the server and losing the opportunity to log more detailed information about which component caused the timeout.
In this change, I subtract 100ms from the deadline on the server side of our interceptors, using our existing serverInterceptor. I also check that there is at least 100ms remaining in which to do useful work, so the server doesn't begin a potentially expensive task only to abort it.
Fixes#3608.
This PR updates the `test/boulder-tools/tag_and_upload.sh` script to template a `Dockerfile` for building multiple copies of `boulder-tools`: one per supported Go version. Unfortunately this is required because only Docker 17+ supports an env var in a Dockerfile `FROM`. It's best if we can stay on package manger installed versions of Docker which precludes 17+ 😞.
The `docker-compose.yml` is updated to version "3" to allow specifying a `GO_VERSION` env var in the respective Boulder `image` directives. This requires `docker-compose` version 1.10.0+ which in turn requires Docker engine version 1.13.0+. The README is updated to reflect these new requirements. This Docker engine version is commonly available in package managers (e.g. Ubuntu 16.04). A sufficient `docker-compose` version is not, but this is a simple one binary Go application that is easy to update outside of package managers.
The `.travis.yml` config file is updated to set the `GO_VERSION` in the build matrix, allowing build tasks for different Go versions. Since the `docker-compose.yml` now requires `docker-compose` 1.10.0+ the
`.travis.yml` also gains a new `before_install` for setting up a modern `docker-compose` version.
Lastly tools and images are updated to support both Go 1.10 (our current Go version) and Go 1.10.1 (the new point release). By default Go 1.10 is used, we can switch this once staging/prod are updated.
_*TODO*: One thing I haven't implemented yet is a `sed` expression in `tag_and_upload.sh` that updates both `image` lines in `docker-compose.yml` with an up-to-date tag. Putting this up for review while I work on that last creature comfort._
Resolves https://github.com/letsencrypt/boulder/issues/3551
Replaces https://github.com/letsencrypt/boulder/pull/3620 (GH got stuck from a yaml error)
This allows these tools to easily be run in command line mode from
the host machine against a Boulder running inside docker-compose up
without modifying the FAKE_DNS field in docker-compose.yml. This
allows for easier testing of various conditions.
This PR updates the RA such that certificateRequestEvent objects created during issuance and written to the audit log as JSON also include a new Authorizations field. This field is a map of the form map[string]certificateRequestAuthz and can be used to map from an identifier name appearing in the associated certificate to a certificateRequestAuthz object. Each of the certificateRequestAuthz objects holds an authorization ID and the type of challenge that made the authorization valid.
Example Audit log output (with the JSON pulled out and pretty-printed):
{
"ID":"0BjPk94KlxExRRIQ3kslRVSJ68KMaTh416chRvq0wyA",
"Requester":666,
"SerialNumber":"ff699d91cab5bc84f1bc97fc71e4e27abc1a",
"VerifiedFields":["subject.commonName","subjectAltName"],
"CommonName":"rand.44634cbf.xyz",
"Names":["rand.44634cbf.xyz"],
"NotBefore":"2018-03-28T19:50:07Z",
"NotAfter":"2018-06-26T19:50:07Z",
"RequestTime":"2018-03-28T20:50:07.234038859Z",
"ResponseTime":"2018-03-28T20:50:07.278848954Z",
"Authorizations": {
"rand.44634cbf.xyz" : {
"ID":"jGt37Rnvfr0nhZn-wLkxrQxc2HBfV4t6TSraRiGnNBM",
"ChallengeType":"http-01"
}
}
}
Resolves#3253
* Update `globalsign/certlint` to d4a45be.
This commit updates the `github.com/globalsign/certlint` dependency to
the latest tip of master (d4a45be06892f3e664f69892aca79a48df510be0).
Unit tests are confirmed to pass:
```
$ go test ./...
ok github.com/globalsign/certlint 3.816s
ok github.com/globalsign/certlint/asn1 (cached)
? github.com/globalsign/certlint/certdata [no test files]
? github.com/globalsign/certlint/checks [no test files]
? github.com/globalsign/certlint/checks/certificate/aiaissuers [no
test files]
? github.com/globalsign/certlint/checks/certificate/all [no test
files]
? github.com/globalsign/certlint/checks/certificate/basicconstraints
[no test files]
? github.com/globalsign/certlint/checks/certificate/extensions [no
test files]
? github.com/globalsign/certlint/checks/certificate/extkeyusage [no
test files]
ok github.com/globalsign/certlint/checks/certificate/internal
(cached)
? github.com/globalsign/certlint/checks/certificate/issuerdn [no
test files]
? github.com/globalsign/certlint/checks/certificate/keyusage [no
test files]
? github.com/globalsign/certlint/checks/certificate/publickey [no
test files]
? github.com/globalsign/certlint/checks/certificate/publickey/goodkey
[no test files]
ok github.com/globalsign/certlint/checks/certificate/publicsuffix
(cached)
? github.com/globalsign/certlint/checks/certificate/revocation [no
test files]
? github.com/globalsign/certlint/checks/certificate/serialnumber
[no test files]
? github.com/globalsign/certlint/checks/certificate/signaturealgorithm
[no test files]
ok github.com/globalsign/certlint/checks/certificate/subject (cached)
ok github.com/globalsign/certlint/checks/certificate/subjectaltname
(cached)
? github.com/globalsign/certlint/checks/certificate/validity [no
test files]
? github.com/globalsign/certlint/checks/certificate/version [no test
files]
? github.com/globalsign/certlint/checks/certificate/wildcard [no
test files]
? github.com/globalsign/certlint/checks/extensions/adobetimestamp
[no test files]
? github.com/globalsign/certlint/checks/extensions/all [no test
files]
? github.com/globalsign/certlint/checks/extensions/authorityinfoaccess
[no test files]
? github.com/globalsign/certlint/checks/extensions/authoritykeyid
[no test files]
? github.com/globalsign/certlint/checks/extensions/basicconstraints
[no test files]
? github.com/globalsign/certlint/checks/extensions/crldistributionpoints
[no test files]
? github.com/globalsign/certlint/checks/extensions/ct [no test
files]
? github.com/globalsign/certlint/checks/extensions/extkeyusage [no
test files]
? github.com/globalsign/certlint/checks/extensions/keyusage [no test
files]
? github.com/globalsign/certlint/checks/extensions/nameconstraints
[no test files]
ok github.com/globalsign/certlint/checks/extensions/ocspmuststaple
(cached)
? github.com/globalsign/certlint/checks/extensions/ocspnocheck [no
test files]
? github.com/globalsign/certlint/checks/extensions/pdfrevocation
[no test files]
? github.com/globalsign/certlint/checks/extensions/policyidentifiers
[no test files]
? github.com/globalsign/certlint/checks/extensions/smimecapabilities
[no test files]
? github.com/globalsign/certlint/checks/extensions/subjectaltname
[no test files]
? github.com/globalsign/certlint/checks/extensions/subjectkeyid [no
test files]
ok github.com/globalsign/certlint/errors (cached)
? github.com/globalsign/certlint/examples/ct [no test files]
? github.com/globalsign/certlint/examples/specificchecks [no test
files]
```
* Certchecker: Remove OCSP Must Staple err ignore, fix typos.
This commit removes the explicit ignore for OCSP Must Staple errors that
was added when the upstream `certlint` package didn't understand that
PKIX extension. That problem was resolved and so we can remove the
ignore from `cert-checker`.
This commit also fixes two typos that were fixed upstream and needed to
be reflected in expected error messages in the `certlint` unit test.
* Certchecker: Ignore Certlint CN/SAN == PSL errors.
`globalsign/certlint`, used by `cmd/cert-checker` to vet certs,
improperly flags certificates that have subj CN/SANs equal to a private
entry in the public suffix list as faulty.
This commit adds a regex that will skip errors that match the certlint
PSL error string. Prior to this workaround the addition of a private PSL
entry as a SAN in the `TestCheckCert` test cert fails the test:
```
--- FAIL: TestCheckCert (1.72s)
main_test.go:221: Found unexpected problem 'Certificate subjectAltName
"dev-myqnapcloud.com" equals "dev-myqnapcloud.com" from the public
suffix list'.
```
With the workaround in place, the test passes again.
Also instead of repeating the same bucket definitions everywhere just use a single top level var in the metrics package in order to discourage copy/pasting.
Fixes#3607.
This commit addresses two config elements that were defined but not
wired through to the WFE implementation object. Prior to this commit the
`c.WFE.DirectoryCAAIdentity` and `c.WFE.DirectoryWebsite` configuration
values were read and unmarshaled from config but not passed to the WFE.
After this commit these two config options will be picked up by the WFE
impl.
In `getAuthorizations`, we had a single loop to both select the freshest authz and
fetch challenges corresponding to authzs. This meant that in some cases, we
would fetch challenges only to throw them away. Since each challenge fetch is a
DB round trip, this would cause extreme slowness when called for domains that
have a large number of authorizations.
This change splits that into two loops: One to select the freshest authzs, and
another to fetch challenges for those authzs.
The upstream `certlint` package doesn't understand the RFC 7633 OCSP
Must Staple PKIX Extension and flags its presence as an error. Until
this is resolved upstream this commit updates `cmd/cert-checker` to
ignore the error.
The `TestCheckCert` unit test is updated to add an unsupported extension
and the OCSP must staple extension to its test cert. Only the
unsupported extension should be flagged as a problem.
This commit removes the possibility of unmarshaling the `Agreement`
field of a new account request. This is a legacy V1 concept and has
no bearing on ACMEv2. Instead a specific one-off struct that only
contains the fields we wish to allow update for (Contact and Status) is
used to unmarshal the new account request.
In #3584 we mounted the package path in an attempt to speed things up,
but that wasn't actually effective. The latest Go has a $GOCACHE which
defaults to $HOME/.cache/go-build, and stores cached build artifacts.
This change mounts that path to a hidden directory on the host. It also
removes the volume mounts for /tmp/ and $GOPATH/pkg, which weren't doing
anything.
I've verified this speeds up the startup for RUN=integration on my
machine from several seconds to roughly zero seconds on repeated runs.
This commit updates the WFE2 to remove the "Agreement" value on V2 account objects before returning them to the user. This field is not defined in the V2 specification and we should not be returning it.
The V2 `TermsOfServiceAgreed` field is marked optional, and for Let's Encrypt purposes it doesn't make much sense to write it in returned Account objects because the value will necessarily be true 100% of the time. We never create an account unless the request has `TermsOfServiceAgreed: true`.
Resolves https://github.com/letsencrypt/boulder/issues/3590
The small package local `assertJSONEquals` function was just calling
`test.AssertUnmarshaledEquals`. This commit removes `assertJSONEquals`
and updates all of the callers to just use
`test.AssertUnmarshaledEquals` themselves.
This commit updates the WFE and WFE2 to have configuration support for
setting a value for the `/directory` object's "meta" field's
optional "caaIdentities" and "website" fields. The config-next wfe/wfe2
configuration are updated with values for these fields. Unit tests are
updated to check that they are sent when expected and not otherwise.
Bonus content: The `test.AssertUnmarshaledEquals` function had a bug
where it would consider two inputs equal when the # of keys differed.
This commit also fixes that bug.
This commit updates CTPolicy & the RA to return a distinct error when
the RA is unable to fetch the required SCTs for a certificate when
processing an issuance. This error type is plumbed up to the WFE/WFE2
where the `web/probs.go` code converts it into a server internal error
with a suitable user facing error.
The outer `config.SubscriberAgreementURL` field has been deprecated for
a while in favour of `config.wfe.SubscriberAgreementURL`. After
verifying the prod/staging configurations do not use the legacy field
this commit removes it.
In some places in the WFE/WFE2 we were calling `logEvent.AddError` and
adding a message that was ~= identical to the `detail` of
a `ProblemDetails` object returned through the API. For these cases this
commit removes the `.AddError` call. We can reference the information
from the API level error and this will save us log bytes overall.
This commit maintains instances where we call `logEvent.AddError` to add
a message with *more* detail than is returned through the API (e.g.
including ID #s or internal error strings).
* Reject WFE2 certificate chain PEM files with CRLF endings.
This commit updates the `boulder-wfe2` command's processing of
certificate chains such that it will reject chain files that contain PEM
encoding with Windows CRLF line endings. Boulder is a UNIX service and
throughout we assume UNIX newlines. CRLF endings in a certificate chain
input file is an error that should be resolved by the operator prior to
startup.
* Add trailing newline to PEM chainfiles automatically.
If a PEM encoded chain file doesn't end with a trailing `\n` the WFE2
should add it. This commit updates the chain file loading to handle this
and adds a corresponding unit test.
Account objects contain email addresses, which the subscriber may choose to
change or delete, and which we protect under our privacy policy. Since audit
logs are retained much longer than regular logs, keeping email addresses out of
the audit logs improves the privacy properties of our email storage.
In publisher and in the integration test, check that SCTs are in a
reasonable range. Also, update CreateTestingSignedSCT (used by
ct-test-srv) to produce SCTs correctly with a timetamp in Unix epoch
milliseconds.
- Remove acme-v2 test phase.
- Rename integration-test-v2.py to v2_integration, so it can be imported.
- Import all symbols from v2_integration before running test_*.
- In chisel2:
- Rename DIRECTORY so it doesn't collide.
- Incidental logging and error fixes.
- Merge v1 and v2 load testing into a single function.
- Run cert-checker just once, after all other test cases.
- In v2_integration:
- Remove unnecessary imports.
- Import chisel2 methods in the chisel2 namespace so they don't
collide with chisel methods.
- Remove main and shutdown code.
Prior to this commit an order's expiry was set based on
ra.orderLiftime while pending and valid authorization expiry was set
based on ra.pendingAuthorizationLifetime and
ra.authorizationLifetime. Since orders reused existing valid/pending
authorizations this can lead to a case where an order has an expiry
beyond the associated authorization expiries. In this case when an
authorization expired the order becomes inactionable and the extra order
lifetime is not useful.
This commit addresses this problem in two ways:
1. The SA GetAuthorizations function used to find authzs to reuse for
ra.NewOrder is adjusted to only return authorizations at min 24hr
away from expiry.
2. Order expiry is now calculated by the RA in newOrder
as the min of the order's own expiry or the soonest
authorization expiry. This properly reflects the order's true
lifetime based on the authorization lifetime.
The RA/SA unit tests are updated accordingly.
Resolves#3498
Roland Bracewell Shoemaker