Checksum and .asc files beside a versioned artifact are as immutable as the
artifact, but were routed through the metadata path: uncached until a
first successful fetch, with no retry or circuit breaker. A build pulling
hundreds of artifacts sent as many checksum requests straight to Maven
Central, which rate-limited the proxy, after which every further checksum
request still went upstream and failed.
Only maven-metadata.xml and its own checksums stay on the metadata path,
since the version listing changes with each release.
A version's integrity is a single value shared by every archive cached
for it, but a dev branch caches one archive per commit. Recording the
listed head's shasum made a cache hit for an older pinned commit fail its
check, so its archive was discarded and fetched again, which cleared the
head's recorded shasum in turn. The listed shasum only ever describes the
current head, and Composer checks dist.shasum itself.
- Skip the shasum for dev versions
- Test a pinned commit, then the head, then both again from the cache
npm versions published before dist.integrity existed carry only a SHA-1
dist.shasum, and Composer metadata lists only a SHA-1 dist.shasum, so
neither could be checked.
- Accept native integrity given as an algorithm-prefixed hex checksum,
verified with integrity's exact-digest check, so a recorded "sha1-<hex>"
is also checked on cache hits
- npm: fall back to dist.shasum when dist.integrity is missing or cannot
be parsed; a usable dist.integrity still takes precedence
- Composer: verify an archive against its dist.shasum, except another
commit of a branch fetched in place of the listed one, which the listed
shasum does not describe
- Fetch unverified as before when no shasum is listed, as Packagist does
for GitHub archives, or it is not a SHA-1 digest
The proxy cached the first tarball upstream sent for a version without
checking it against the packument, so a tampered tarball was served to
every client after it.
- Pass the packument's dist.integrity to the existing integrity checks, so
a fresh tarball is verified as it is stored or streamed; a mismatch is
discarded and returns 502, or aborts the streamed response
- Record it with the cached version, so cache hits are verified against it
through the same check as before
- Verify with integrity.Result.Verify, which accepts any digest of the
strongest algorithm listed
- Fetch unverified as before when dist.integrity is missing or cannot be
parsed, or the packument cannot be fetched
Docker sends one Accept header per media type. Reading only the first
forwarded just that type upstream and ruled out cached manifests of the
others, so pulls failed when the upstream was down.
- GetLocalArtifact went through checkCache, which reports a miss when StreamArtifacts is set, so npm and Composer downloads fetched the upstream archive instead of the pushed one
- Split readStored out of lookupCache and read pushed artifacts through it regardless of StreamArtifacts
- Keep the denylist check and the 503/404 answers for unreadable or missing pushed artifacts
- Test npm and Composer downloads of pushed archives with StreamArtifacts set
Refs #133
Pushing and downloading refuse denied versions, but npm and Composer metadata listed every pushed version. A version denied after it was pushed stayed listed, so resolvers picked it and then got 403 on the download. localVersions now leaves denied versions out, which covers both ecosystems and the hosted-only namespaces; npm latest falls back to the highest allowed version. /api/hosted still lists them, so operators can see and delete them.
Refs #134
The manifest lookup took the first composer.json at the root or inside a top directory, in zip order. Dist archives keep composer.json files for tests and examples, so tests/composer.json stored before the root file refused the push with a name mismatch. The root composer.json now always wins. Without one, the manifest inside a top directory is used only when there is exactly one; more than one refuses the archive as ambiguous.
Refs #134
checkCache answered nil both when a stored object couldn't be opened and after an integrity failure had cleared the record. Download handlers then fetched the upstream archive for the same version and filename, and recording it turned the pushed row into an upstream one without its manifest and queued the pushed object for deletion. One storage error while downloading a patched version served the unpatched upstream archive and dropped the push.
checkCache now looks at the record on a miss. For a pushed artifact it returns ErrHostedArtifactUnavailable when the object is recorded but couldn't be served, and ErrHostedArtifactMissing when no object is recorded any more. Every path that would fetch upstream goes through it, so none of them can replace the pushed artifact. Downloads answer 503 and 404 for the two.
Refs #133
A package the upstream doesn't know is served from its pushed versions, but the upstream is still asked about it. Once someone publishes the same name there, their versions are listed next to the pushed ones and a version range can resolve to them.
hosted.npm_scopes and hosted.composer_vendors (PROXY_HOSTED_NPM_SCOPES, PROXY_HOSTED_COMPOSER_VENDORS) name namespaces whose packages are served from pushed versions alone. Their metadata is built from what was pushed, npm latest included, and anything not pushed answers 404. The upstream is never asked about them. Packages in other namespaces are proxied as before.
The npm metadata handler's decisions about pushed versions move into serveLocalMetadata.
Refs #133
Refs #134
The packument lists pushed versions without cooldown, but the tarball download checked cooldown against the upstream's publish times before looking at the cache. A pushed override of a version the upstream published recently was listed and then withheld with 404. Pushed tarballs are now served right after the denylist check.
Refs #134
With hosted packages enabled, a package's p2 metadata lists its pushed versions, built from composer.json without the root-only keys, with a zip dist at the proxy and the time they were pushed. dist.reference is the archive's sha256 so a re-push isn't served from Composer's cache, and shasum is left empty. Dev versions go to the ~dev file and tagged versions to the other.
A pushed version takes the place of an upstream entry with the same version, with or without a leading v. Pushed entries follow the upstream ones and go through the same minified writer, so fields an earlier upstream entry set and the pushed one lacks are written as __unset rather than inherited.
When the upstream doesn't know the package, the document holds only the pushed versions. Like npm, documents with pushed versions bypass the rewrite cache.
Refs #134
With hosted packages enabled, the packument of a package with pushed versions lists each of them, built from its package.json with dist pointing at the proxy and carrying the archive's sha512, and the time it was pushed. A pushed version takes the place of an upstream entry with the same version. Cooldown and the denylist still filter only upstream versions, and dist-tags stay as upstream has them.
When the upstream doesn't know the package, the packument holds only the pushed versions, with latest set to the highest of them. Requests for a single pushed version are answered from its entry.
Rewrites that include pushed versions bypass the rewrite cache, which is keyed by the upstream document alone, so a push or delete shows on the next request and changes the ETag. Packages without pushed versions take the cached path as before, and with hosted packages disabled there is no extra lookup.
Refs #134
Resolvers take a version's dependencies from the registry metadata, not from the archive, so listing a pushed version needs its manifest. A push now reads package.json from the top directory of an npm tarball, or composer.json from the root or the single top directory of a Composer zip, and keeps it in the new artifacts.manifest column (migration 013_add_artifact_manifest).
The manifest has to name the pushed package and version. A Composer manifest may leave the version out, and its name is compared case-insensitively. An archive that can't be read or whose manifest names something else is refused with 400 and removed from storage.
The version's integrity is now the archive's sha512, which npm clients expect in dist.integrity.
Refs #134
With hosted.enabled, PUT /api/hosted/{ecosystem}/{name}/{version} stores the request body as the archive for that version, marked local. It replaces any cached upstream copy and is served from the cache without contacting the upstream. GET /api/hosted/{ecosystem}/{name} lists what was pushed and DELETE removes a version again, after which clients get the upstream archive if there is one. Pushing and deleting need the bearer token; listing is open like the rest of the API.
Download handlers look archives up by the filename clients request, so a push is stored under the name its ecosystem derives from name and version: {name}-{version}.tgz for npm, without the scope, and {package}-{version}.zip for Composer. Other ecosystems are refused for now.
The Composer download handler resolves the upstream metadata before it looks at the cache and answers 404 for versions the upstream doesn't list. It now serves a pushed archive first. Only artifacts marked local take that path, so cached upstream archives still go through the metadata and commit checks.
The version's integrity is set to the pushed bytes, so the cache's integrity check verifies them rather than an upstream digest. Names and versions with path separators, dot segments or control characters are refused before anything is stored.
Refs #133
Artifacts had no size limit. A fetch was bounded only by the server's write
timeout, and storage.max_size evicts only after the bytes are written, so
any client could fill storage or keep fetches busy by requesting huge
artifacts.
- Add artifact_max_size (PROXY_ARTIFACT_MAX_SIZE), unlimited by default
- Refuse an artifact whose declared size is over the limit before reading
it, and fail one of unknown size once the download passes the limit
- Apply it in the fetcher so stored, streamed and mirrored artifacts all
honour it; nothing is cached and clients get 502
Handlers set the configured upstream auth header on the request itself
before handing it to the client. When upstream redirected to another host,
Go's client copied that header to the next hop: it strips only
Authorization and Cookie, so a "header" type credential such as
X-JFrog-Art-Api reached the redirect target. The auth transport already
applies credentials per hop, scoped to each hop's URL.
- Remove Proxy.applyUpstreamAuth and Proxy.AuthForURL; auth comes only
from the transport
- Add a server test that redirects a header-authenticated upstream to
another host and asserts the header stays with the upstream
- Wire handler tests' auth through the transport, as the server does
Cooldown only filtered the metadata. A composer.lock names the proxy's
/files URL directly, so `composer install` fetched versions still inside
the cooldown window without ever seeing the filtered metadata. npm already
checks the publish time on tarball downloads.
- Return the version's publish time with its dist URL, following the
minified format's inheritance
- Answer 404 for a version in cooldown before fetching its archive
- Read the download lookup's metadata through ReadMetadata, so it has the
same size limit as other metadata reads
Packages are stored as pkg:deb/{name}, without a vendor and whichever configured archive they came from. Package rules may name a vendor the way PURLs usually do (pkg:deb/debian/curl, pkg:deb/ubuntu/curl); it's dropped, so a rule covers the package from every archive. The docs row warns that Debian and Ubuntu archives only keep the current version per suite.
The mirror can't fetch deb packages, so there's no mirror test.
Refs #306
On a build network where clients cannot reach sum.golang.org at all, the
proxy should keep answering for the database during an outage, since a 404
on supported would only send the go command somewhere it cannot go. With
the switch set, a cached tree of any age is enough; without it, the tree
has to be fresh, as before.
That does not make every cached module verify. The go command checks a
lookup against tiles of the tree that came with it, and sum.golang.org
attaches a different tree to almost every lookup, so it can ask for a tile
nobody fetched through the proxy before. configuration.md says so.
While the database is known to be unreachable, offline mode also serves
cached lookups and tiles straight away. Otherwise each of them would sit
through the upstream timeout before the fetch falls back to the cache,
and a build would take minutes for as long as the outage lasts. A lookup
or tile that falls back to the cache renews what the proxy knows about
the database, unless a probe has reached it since the request started.
The go command asks supported only once per process, so a long build would
otherwise be back to waiting a minute into the outage. recordTimeout is
now recordFailure, since lookups use it too.
The option needs cache_metadata, since without it there is no cached tree
to fall back on, and the config refuses the combination.
Also the cases around a silent or slow database: a fresh cached tree still
gets a yes on supported, a stale one does not, a late answer from the
database clears the remembered failure, a fallback to the cached tree is
not mistaken for an answer, and an answer still counts when the cache
write fails afterwards. The damaged tile test guards the size check on the
cached copy of a complete tile: it shortens the stored blob and expects
the tile to be fetched again, also past the TTL from a database that sends
an ETag.
A block page in place of a lookup must not replace the cached record, not
even one that starts like a record. The lookup fixtures now look like real
records, since lookups are checked too.
Two problems with the check behind supported. The fetch it waits on is
shared with other callers and only bounded by the upstream timeout, which
defaults to 30 seconds. A firewall that silently drops packets to
sum.golang.org would stall the first go command every minute for that long,
and the go command itself has no timeout on this request. Now we wait five
seconds at most; the fetch carries on in the background, records how it
went and fills the cache for the next check. A timeout only counts as a
failure if the fetch hasn't succeeded in the meantime.
The fetch reports no error when it falls back to the cached tree after the
database failed, so the metadata result now says when it did, and the
check treats that as the database not answering. A fresh cached tree
(within metadata_ttl) still lets a timed-out check say yes, since the
fetch only falls back once upstream has failed, which a silent database
delays past our timeout. An older tree does not count: once the proxy has
lost the database, clients go there directly, as they would without the
proxy.
And the check believed any 200. Egress proxies tend to answer blocked hosts
with a 200 HTML page, which we would have cached as the signed tree and taken
as proof that the database is reachable. latest now has to start with
"go.sum database tree", or it is refused like a bad tile.
Lookups had the same gap. A block page cached as a lookup would also
replace the good copy that is served during an outage, so a lookup now has
to be a record followed by a signed tree, the way the go command reads it.
Complete tiles are never fetched again once cached, so a bad response would
stick around until someone deletes it by hand. That could be an error page
from a proxy in between that comes back with 200.
A hash tile is always W*32 bytes, so anything else is refused now and not
stored, and the same check runs when a cached tile is served. Data tiles
have no fixed size and only must not be empty.
A cached tile that fails the check is fetched again without asking the
database whether it changed. Past the TTL the cache would ask, and a
database that sends an ETag would answer 304 and hand back the damaged
copy every time.
The fake database in the tests now guards its maps with the mutex, since the
tests change them between requests.
Once a proxy answers 200 on supported, the go command sends all checksum
requests there and never goes to the database directly. A proxy that cannot
reach sum.golang.org would therefore break every build that needs a lookup.
supported now fetches latest first, through the cache, and answers 404 if
that fails and nothing is cached, so clients fall back like before. A failed
check is remembered for a minute so an unreachable database doesn't make
each go command wait for an upstream timeout.
Everything under /go/sumdb/ was passed on to the module upstream. But
proxy.golang.org doesn't serve the checksum database, so those requests
always came back 404 and every client had to talk to sum.golang.org itself.
The Go handler now serves /sumdb/{name}/ for the databases in
upstream.go_sumdbs, which defaults to sum.golang.org. Lookups, latest and
tiles come from the database through the metadata cache. Complete tiles
never change, so once they're cached we don't ask again. Unknown names and
paths that aren't part of the protocol get a 404 without going upstream.
If the database answers 404 we pass on up to 4 KiB of its text, because the
go command prints it and it tells you why a version is missing. For that the
not-found error from the metadata fetch now carries the start of the body.
Other handlers still just see ErrUpstreamNotFound.
The first matching pattern names the package, so an upstream's own
patterns now take precedence over overlapping built-ins and keep the
package names, cache entries and rules already keyed on them.
