| Age | Commit message (Collapse) | Author |
|
Fixes golang/go#74027.
Change-Id: I916ac81093e782d4eda21fe11ef47eeff4f5f0b1
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/751480
Reviewed-by: Jonathan Amsterdam <jba@google.com>
kokoro-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ethan Lee <ethanalee@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
|
|
- In preparation for upgrade to PostgreSQL 17, upgrade stale
dependencies within pkgsite.
Change-Id: Id54cbb46182711b9808ee3bbbe2f0d5d9a04583d
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/758121
Auto-Submit: Ethan Lee <ethanalee@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Jonathan Amsterdam <jba@google.com>
kokoro-CI: kokoro <noreply+kokoro@google.com>
|
|
Currently, InMemory queue sits in queue, and is instantiated in gcpqueue for
convenience. In preparation for a third queue type (Postgres), this CL
separates the two more cleanly, making it more ergonomic for the new queue type
to slot in next to the existing two.
This CL doesn't change any logic: it just exists to make the next CL smaller
and easier to review.
I also took the liberty of adding some tests specific to the InMemory queue,
since I didn't find any. Let me know if it's tested in another place, though,
and if you'd prefer me to remove it.
Updates golang/go#74027.
Change-Id: I44bd92129f33bc7975fcd138c905e0b7ab49d257
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/747881
kokoro-CI: kokoro <noreply+kokoro@google.com>
Auto-Submit: Jonathan Amsterdam <jba@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ethan Lee <ethanalee@google.com>
Reviewed-by: Jonathan Amsterdam <jba@google.com>
|
|
Fixes golang/go#77689.
Change-Id: Id754eeafdc8b680acb1aa137225292c808d0bf04
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/747880
Reviewed-by: Jonathan Amsterdam <jba@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
kokoro-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: David Chase <drchase@google.com>
|
|
The beta environment is no longer in use. This change removes related
logic and references, such as the middleware redirect to
beta.pkg.go.dev.
Change-Id: I0c6ebbe11d5689ccf2abf8c722399d8dc339db89
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/715321
kokoro-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jonathan Amsterdam <jba@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Matloob <matloob@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
|
|
This CL provides several improvements to the worker home page:
- Display all requests, not just fetches.
- Link to the logs for each request.
- Link that will cancel a request.
At the heart of these changes is a new piece of middleware that tracks
all active requests, along with their trace ID and a function that
can be used to cancel them.
This change also affects logging, because the logger doesn't need to
maintain its own trace ID.
Change-Id: Id022170073d2d7ca4e45aaa1d78b216d8a512f35
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/568236
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
|
|
- Move excluded prefix list to a separate page.
- Move debug pages under "/debug".
(Previously the dcensus ones weren't even exposed.)
- Add links to debug pages at top of home page.
Change-Id: If26e139984b34cb6b2ef78af8ef154d96436a26a
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/566418
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
kokoro-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
|
|
This package adds a StartSpan function that does nothing unless an
implementation for the StartSpan function is injected in. That allows
us to have the frontend and worker mains inject the implementation
while internal/fetch and internal/source can use the StartSpan
function without depending on opencensus's tracing library.
For golang/go#61399
Change-Id: I2c650bdbb56c8fcc47ffabab8353cef74e0d1c4e
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/523739
Reviewed-by: Robert Findley <rfindley@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Run-TryBot: Michael Matloob <matloob@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
|
|
This removes the dependency from internal/proxy on ochttp.
For golang/go#61399
Change-Id: Ib2a350151752dc18c9a6e03b38b2d423ed55f4a1
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/523513
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Run-TryBot: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
|
|
This removes the dependency from package source onto ochttp. The users
of source.NewClient that want an ochttp.Transport can set the
transport on the *http.Client.
For golang/go#61399
Change-Id: Ifb7126c482f664ee5a359f594d9324f0fd90f8b2
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/523510
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
|
|
This change creates a new package that does config initialization and
other GCP-specific operations that were previously done in package
config, so that config can have no cloud dependencies.
For golang/go#61399
Change-Id: I8d78294834e325b47d838892a1cef87003a4b90a
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/522516
Run-TryBot: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
kokoro-CI: kokoro <noreply+kokoro@google.com>
|
|
This removes the dependency of internal/frontend on the cloud error
reporting client, both directly, and through the derrors package by
introducing a new interface Reporter that is used both to set the
reporting client for internal/derrors, and on the Server.
For golang/go#61399
Change-Id: Id4d4def522cda9b4e49f53cff6708019dec2693c
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/514676
Reviewed-by: Jamal Carvalho <jamal@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
|
|
This removes the direct dependency from cmd/pkgsite to the middleware
package. The middleware package has a lot of dependencies but most
middlewares are not used by cmd/pkgsite.
There are still some middlewares used by internal/frontend so the
indirect dependency still exists.
For #61399
Change-Id: I09eee46b2ff13e112d24df62073074eb5aa3e901
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/511315
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
|
|
internal/frontend depends on internal/queue. Move the parts with gcp
dependencies into its own package so internal/frontend no longer has
those gcp queue dependencies.
For #61399
Change-Id: I0083c31da1b367f135e53686e4c994d7c4d6420f
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/510815
TryBot-Result: Gopher Robot <gobot@golang.org>
kokoro-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
|
|
For golang/go#59622.
Change-Id: I4059df405e248824a3521fcc084568b052fe4380
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/486458
Reviewed-by: Suzy Mueller <suzmue@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Jamal Carvalho <jamal@golang.org>
|
|
Added an endpoint to clear the cache during cloudbuild
deployments of the beta frontend.
Change-Id: I3fb19c3f1eef6b1b15eda2601b27a403b39b0f8d
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/375956
Trust: Jamal Carvalho <jamal@golang.org>
Run-TryBot: Jamal Carvalho <jamal@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Add metrics for the numbers of active and waiting DB processes.
For golang/go#48010
Change-Id: Ia3c14e492b29c07371ee903182c7ba55f04c584a
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/349310
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Fetch latency and related metrics include the time spent inserting
into the DB, as well as the time to fetch and process the module.
For golang/go#48010
Change-Id: I1d685bd25f1b632b0b20de5b1bfac5003bff0caa
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/346750
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Move the load-shedding logic to the worker and have it span both the
fetch and processing of the module (as previously) as well as
inserting it into the database.
This is a more accurate estimation of load, since running a lot of
concurrent queries definitely slows down processing.
Most of the time this won't make much difference, but under high load,
such as when processing multiple large modules, it will reduce DB
contention and should result in greater throughput.
For golang/go#48010
Change-Id: I7d0922e02d00182e867fd3b29fc284c32ecab5ee
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/346749
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Change-Id: Ie5d3e056fb2507d2e0e4b093f4a437031841ed55
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/342672
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
We no longer need a special list of removed modules, because the
disable-fetch header serves the same purpose.
Change-Id: Ia52a5cfc1036127b78d36f668917809952c57e81
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/341532
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
The redis HA was used by search autocomplete and is no longer used by
anything.
Change-Id: I39a2c81c47da53823d7b668a980159c4726c6b13
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/340393
Trust: Julie Qiu <julie@golang.org>
Run-TryBot: Julie Qiu <julie@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jonathan Amsterdam <jba@google.com>
|
|
Factor out this commonly used function to the internal package.
Also, update some comments in devtools/cmd/seeddb.
Change-Id: I0276ce51d0dd1f4aafa2f6c562d1b83ce21a0d14
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/337330
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
log.SetLevel is now called in config.Init, so that it also applies to
logs in the Init function.
Change-Id: Id552923b4d81bb1678b8ffa72f1b1d858de559c1
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/335257
Trust: Julie Qiu <julie@golang.org>
Run-TryBot: Julie Qiu <julie@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
|
|
The pkgsite logger is now used instead of the log package from the
standard library.
Change-Id: I5083d8cd6bf7a96a9245f848633b8803d0d80483
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/333931
Trust: Julie Qiu <julie@golang.org>
Run-TryBot: Julie Qiu <julie@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jonathan Amsterdam <jba@google.com>
|
|
Change-Id: I41990fe0fe344ee8038e5841a242b087e3065540
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/327751
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Change-Id: Icec167d645b9178375b959491206277e13b455cf
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/327283
Trust: Jamal Carvalho <jamal@golang.org>
Run-TryBot: Jamal Carvalho <jamal@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Add a metric that records how many modules need to be processed (those
with status codes of 0 or 5xx). That will tell us how much the worker
is backed up.
Change-Id: I68c47255e3720b5f978037c1251b0e502961d089
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/323109
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Change-Id: Id211af73d327d92d3cf361b5909619ac14856860
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/298751
Trust: Jamal Carvalho <jamal@golang.org>
Run-TryBot: Jamal Carvalho <jamal@golang.org>
Reviewed-by: Julie Qiu <julie@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
|
|
Convert the proxy client to one which doesn't fetch at the beginning
of the request, so we don't have to pass a boolean argument through
several layers.
For golang/go#44437
Change-Id: I0495b5673cc399a1cdc16349d53ba4a9a4ecc380
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/295889
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
This directory is created by npm and is large. In fact, including
it results in too many files for `gcloud app deploy`.
Also, put .gcloudignore files under version control.
Change-Id: I6cbf1c42b312911b9b2b16215469cb027bedebe1
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/286932
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
If the fetch handler sees the "proxyfetch" query param
set to "off", it tells internal/fetch to disable
fetching by the proxy.
Change-Id: I1c9303f96f9249a5b420a73451551aba21d9d417
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/278953
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
|
|
Move some common args to a struct.
Change-Id: I4e500bc72f0cc622ec7641568beb5c7dbde1e69a
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/278713
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jamal Carvalho <jamal@golang.org>
|
|
Create a metric for the age of the oldest unprocessed module, and record it periodically.
We choose to record it whenever the /poll endpoint is hit. Since that
endpoint is hit periodically (once a minute) by Cloud Scheduler, we
can be sure that we will get exactly one data point per minute. The
downside is that we won't have finer granularity than that, but since
the main purpose of this metric is to detect large lags, that doesn't
matter.
Change-Id: I7ce3f043cceaf567fb0c871676063eda91bb2420
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/278614
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
|
|
The worker's /fetch endpoint now replies on success with a message
that uses the resolved version of the module, not the requested one.
Change-Id: Id48e624bb9d34a1a1fd2541c1b11956850faf14c
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/274955
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Our previous quota middleware stored data in RAM. That meant that it
was ineffective when there were multiple instances (processes)
running, because the totals were separate.
This CL uses the same redis instance we use for caching
to hold the data, so it's shared.
Also, previously the only way to set some quota parameters was via the
override file, which is intended only for emergencies. Allow the
important ones to be set by environment variables, like most of our
other config.
Also, the rate-limiter package we use requires v8 of the redis client,
so move all other uses of it to v8.
Change-Id: I1804014030db834a30511786f0a7edfd02e279a5
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/269737
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Change-Id: If9be542ae7b50789c548f19dbab419a2db74e8ca
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/272793
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
The worker and frontend get the driver name from the config struct,
via an environment variable.
The default is "postgres", which was the previous hardcoded value, so
this CL on its own has no effect.
Change-Id: Ia1ecd39e15ea75f21dc7c94aa5567b870a646bf7
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/272787
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
The task names we generated for Cloud Tasks included time information,
so that we could control how freqeuently tasks were de-duped. We've
been using a 3-hour window, meaning that if a module's task finished
or was deleted, another task for that module couldn't be added to the
queue (would be de-duped) for 3 hours.
This scheme has the unintended consequence that if a module is still
on the queue after 3 hours, then another task for the same module
could be enqueued (since its task ID would be different). That means
that if the queue ever gets 3 hours behind, it could get filled with
duplicate modules. For example, this morning the queue had about 25,000
tasks, of which about 7,000 were duplicates.
This CL sets a task's name to a function of the module path and
version. That will prevent a task on the queue from being duplicated.
A repeat task can be added about 1 hour after the previous task
finishes or is deleted (see the Task De-duplication section of
https://cloud.google.com/tasks/docs/reference/rpc/google.cloud.tasks.v2#task),
which is fine. It is still possible to add a suffix to the task
name to requeue sooner.
Change-Id: I34ffce5ea67b9e00b88ca4cf38182e34a6ba8657
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/270337
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Double the timeout for writing to the redis cache instance, from 3 to 6 seconds.
We saw an error on /clear-cache that was due to hitting the write timeout:
https://pantheon.corp.google.com/errors/COmQ7oS0hITl3gE?time=P30D&project=go-discovery
https://pantheon.corp.google.com/logs/query;query=resource.type%3D%22gae_app%22%20resource.labels.module_id%3D%22etl%22%0A%0Atrace%3D%22d3d2cbb054cf3b8e4627e049d3c5cc9e%2F8803268222527035810%3Bo%3D1%22;timeRange=2020-10-12T17:57:33.989Z%2F2020-10-13T12:57:33.989Z?project=go-discovery
Change-Id: Ib9aae6ccece58ecccf78b2627c65a176295199ee
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/261817
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
All code paths get experiments from a config file.
Change-Id: I53e3c22e443573546d8ebdf65303012f845b980c
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/259840
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
This should be a smoother metric than the number of fetched modules,
since processing time per package should have a much smaller range
than time per module.
Change-Id: I00f5ba83fcc4fe3c606883bd644e570b502c9090
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/257243
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Add metrics that count the number of shedded and actual fetches,
along with the time of actual fetches.
Also factor out some common code to internal/dcensus.
Change-Id: Icb52fb0d0456ac760ac9d846aea4a77dd0d16c47
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256764
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
|
|
Have config.GetEnvInt terminate the program if the environment
variable's value is non-empty but not an integer. That's clearly
a bug.
Also, use it in more places.
Change-Id: Ibcc192853e4b5fc01dad709de4de2e5db6fff709
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256760
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Validate the header that comes from the IAP. This ensures that
no one is accessing the server in a way that bypasses the IAP.
The worker will perform this validation, but only if
an env var is set, so it won't affect local testing and
we can turn it on or off easily.
Change-Id: I5932548d8a4ef0502f0c936b394f711ac310bcd3
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256759
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
To get the list of experiments, pass in an ExperimentGetter instead of
a DB client.
Change-Id: I89ccd402a178d195c85efcdaf363e59c45358b2f
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256381
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
TryBot-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
If the GO_DISCOVERY_EXPERIMENTS_FROM_CONFIG env var is "true", read
experiments from dynamic config instead of the database.
Temporary, until we switch over completely to dynamic config.
For golang/go#41483
Change-Id: Icf6038fb9f2322eba24bedf60450cd9fce84bd6c
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256380
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
All our ExperimentSources had the same GetExperiment behavior over
time, so the ExperimentSource was an unnecessary level of
indirection. Remove it.
Change-Id: I975d3095d87d074736e48487bc9f8d8b1c9a14f1
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256379
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Change-Id: I0d4bb1c4aafa7ba7c00f5f15cedb475022c1cf86
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256378
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
|
|
Factor out some common configuration code from the worker and frontend
commands.
Change-Id: I640a90ad56f0866f2bdbb621f64de93ec75dfd47
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/256377
Trust: Jonathan Amsterdam <jba@google.com>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Julie Qiu <julie@golang.org>
TryBot-Result: kokoro <noreply+kokoro@google.com>
|