Since issue #660 showed that we could not rely on Basic Auth for some
very particular cases (of POST requests), we wanted to replace Basic
Auth with a pre-generated token based authentication mechanism for these
cases.
However, as there already is such a mechanism used to authenticate
websocket connections, we simply extended the use of that mechanism.
Therefore, the token that is initially generated for websocket
connections is also now a valid token to authenticate any other
connection. The relevant types and method names have been changed to
reflect that. The javascript code pertaining to the sensitive cases
mentioned above has been changed to use that token.
Some doc has been added to point out how security-sensitive the
OpDiscovery permission is, since it gives access to the auth token.
Fixes issue #660
Change-Id: Iafed3b6e4804364ca2559414c8d87dc4a30f6637
Version in binaries will now be:
"X.Y (git rev xxxxx)"
if X.Y is specified with --tarball_version,
"xxxx"
otherwise.
Also uniformized the flags names and use in other programs.
Fixes#665
Change-Id: I958094f69b279437504854a3ff10e924d42c4fb6
Because it can happen that we end up with with an already stopped/killed
demoblob3179, in which case we need to remove it even when killing it is
unsuccessful (because it was already dead).
Change-Id: I289e96284deebeaaa775c23ef6f3fe3bffc3777c
Most of it replaced with vendor/go4.org/types and
vendor/go4.org/readerutil
u32 went where needed in pkg/blobserver/*
invertedBool went in pkg/types/serverconfig
atomics64 went in pkg/fs
Change-Id: I230426cda35be4b45ed67e869f14e6fdae89be22
This change allows (for example) to hook a blobserver and an indexer
together in a way similar as they would be in a default Camlistore setup.
This is particularly useful to setup tests that require a realistic
indexer, without having to create a heavier test with e.g. test.World.
This actual goal of this change is described below. I kept these changes
together because the one described below illustrates the need for the
synchandler changes.
pkg/index/stress/: stress test the indexer with more data
The goal of these benchmkarks is to help us compare the various sorted
implementations, so we can pick the most efficient one as a default.
A second, harder, goal would be to add tests that help us estimate the
reliability of the various kv stores, again so we can rule out the buggy
ones, or help fix them.
Change-Id: I8a1fe50398a4ba09c03786502b68b6c2599e5984
- add an Attrs map to PermanodeMap that is kept in sync with Claims
- update Attrs map on scanFromStorage and whenever a new blobRef is added
- add fast path to AppendPermanodeAttrValues(Locked),
PermanodeAttrValueLocked and PermanodeHasAttrValueLocked
- fix bugs in and add tests for PermanodeHasAttrValueLocked and
PermanodeAttrValueLocked slow path
Change-Id: Ib0dcdcf9355dd5fa8588c872721a029ec114cb6b
Fixes issue #675
Since camput does not seem affected by the above issue, and that camput
(pkg/client really) main difference in how it uploads is that it does
gated requests, it seemed to follow that gating the requests might fix
the problem.
With this change, I don't see the problem anymore either in
TestS3WriteFiles or by manually uploading with the web UI.
I suppose this indicates that S3 throttles incoming connections pretty
strictly, but I have no formal evidence of it.
Change-Id: I48f10923a710f4a21e44bc251e648db37ae504ec
And try to underscore that the last byte is an LF, which must be kept
and accounted for.
Andy Song raised my attention to this ambiguity.
Change-Id: Ia352980350e44b57f47394841f643a737b91dda7
Otherwise when openOrCachedDB is called for the second time,
it would return an actually closed DB.
Change-Id: I820834e508fed1a34ab7bd19a8a91e398badff9c
Because orchardup/mysql is deprecated.
I had to increase the timeout, because this image takes at least twice
as long to start compared with the orchardup one. (tests went from ~5s to
~11s here).
Change-Id: I42c485d736c65ce9d93aabf02b1de73e6ac799ff
Brad Fitzpatrick