david-sarah [Thu, 31 May 2012 22:00:00 +0000 (22:00 +0000)]
Change 'setup.py test' and 'setup.py trial' to pass --rterrors to trial by default. Suppress using --no-rterrors. Also pass --until-failure/-u to trial. fixes #1699
david-sarah [Sun, 20 May 2012 15:35:29 +0000 (15:35 +0000)]
Catch exceptions from CLI in order to prevent the Ubuntu crash monolog from triggering. refs #1746
Brian Warner [Wed, 30 May 2012 07:17:55 +0000 (00:17 -0700)]
node.py: stop stripping whitespace in write_private_config()
It's nice to add newlines to the saved file, so 'cat' is easy to use. We
still strip on the input side, in get_or_create_private_config().
Brian Warner [Tue, 22 May 2012 22:18:26 +0000 (15:18 -0700)]
test_web.py: fix memory leak when run with --until-failure
The Fake*Node classes in test/common.py were accumulating share data in
a class-level dictionary, which persisted from one test run to the next.
As a result, running test_web.py over and over (with trial's
--until-failure feature) made this dictionary grow without bound,
eventually running out of memory.
This fix moves that dictionary into the FakeClient built fresh for each
test, so it doesn't build up. It does the same thing for "file_types",
which was much smaller but still lived at the class level.
Closes #1729
Brian Warner [Tue, 22 May 2012 15:50:36 +0000 (08:50 -0700)]
test/check_memory.py: oops, fix one last ur.uri -> ur.get_uri()
Brian Warner [Tue, 22 May 2012 04:18:37 +0000 (21:18 -0700)]
change UploadResults to return IServers, update users to match
This finally changes all callers of get_servermap()/get_sharemap() to
accept IServers, and changes UploadResults to provide them.
Brian Warner [Tue, 22 May 2012 04:18:25 +0000 (21:18 -0700)]
UploadResults: store IServers internally, but still return serverids
This stores IDisplayableServer-providing instances (StubServers or
NativeStorageServers) in the .servermap and .sharemap dictionaries. But
get_servermap()/get_sharemap() still return data structures with
serverids, not IServers, by translating their data on the way out. This
lets us put off changing the callers for a little bit longer.
Brian Warner [Tue, 22 May 2012 04:17:27 +0000 (21:17 -0700)]
split IDisplayableServer from IServer, add sb.get_stub_server()
IDisplayableServer includes just enough functionality to call
.get_name() and friends, which is all that the UploadResults really
need. IServer is a superset that includes actual share-manipulation
methods. StubServer instances provide only IDisplayableServer, while
actual NativeStorageServer instances provide the full IServer interface.
When the Helper sends a serverid (so we know what to call the server but
nothing else about it, and have no corresponding NativeStorageServer
object to reference), but we want to store an IDisplayableServer in the
UploadResults, we create a synthetic StubServer "server" and store that
instead.
Brian Warner [Tue, 22 May 2012 04:14:44 +0000 (21:14 -0700)]
switch UploadResults to use get_uri(), hide internal ._uri
Complete the getter-based transformation, by hiding ".uri" and updating
callers to use get_uri(). Also don't set a dummy self._uri, leave it
undefined until someone calls set_uri().
Brian Warner [Tue, 22 May 2012 04:14:28 +0000 (21:14 -0700)]
switch UploadResults to use getters, hide internal data, for all but .uri
This hides attributes with e.g. _sharemap, and creates getters like
get_sharemap() to access them, for every field except .uri . This will
make it easier to modify the internal representation of .sharemap
without requiring callers to adjust quite yet.
".uri" has so many users that it seemed better to update it in a
subsequent patch.
Brian Warner [Tue, 22 May 2012 04:14:14 +0000 (21:14 -0700)]
convert UploadResults to a fat init
Populate most of UploadResults (except .uri, which is learned later when
using a Helper) in the constructor, instead of allowing creators to
write to attributes later. This will help isolate the fields that we
want to change to use IServers.
Brian Warner [Tue, 22 May 2012 04:14:00 +0000 (21:14 -0700)]
add HelperUploadResults
This splits the pb.Copyable on-wire object (HelperUploadResults) out
from the local results object (UploadResults). To maintain compatibility
with older Helpers, we have to leave pb.Copyable classes alone and
unmodified, but we want to change UploadResults to use IServers instead
of serverids. So by using a different class on the wire, and translating
to/from it on either end, we can accomplish both.
Brian Warner [Tue, 22 May 2012 04:13:47 +0000 (21:13 -0700)]
Uploader cleanup: create results at end, not beginning
This will make it easier to populate the UploadResults during __init__,
instead of doing it one-field-at-a-time later.
Brian Warner [Tue, 22 May 2012 04:13:32 +0000 (21:13 -0700)]
clean up Helper to make later changes easier
Fix up control flow inside the Helper, to make it more friendly for
later refactoring.
Brian Warner [Tue, 22 May 2012 04:13:11 +0000 (21:13 -0700)]
helper: remove timings["existence_check"], aka "Already-In-Grid Check"
This measured how long the Helper took to do a filecheck before asking
for ciphertext. The "Contacting Helper" report includes both
existence_check and the client-helper RTT.
For non-overlapping uploads, it was being returned correctly. But when
multiple upload requests overlapped, and the file was not already in the
grid, the filecheck would only run once, and its existence_check time
would be reported for all uploaders (even if they didn't have to wait
for that time). Cleaning that up proved too difficult: the only correct
place to report this time is from the initial remote_upload_chk() call,
but the return value of that is too constrained to accomodate it in the
needs-upload case.
So I'm removing it altogether. Eventually I plan to add a proper
events/times field and record more data, including this check, in a form
that can be drawn on a nice zoomable timeline view.
Old clients talking to a new Helper (which doesn't supply the value)
will tolerate the loss (they'll just display an empty field on the web
view).
Brian Warner [Thu, 17 May 2012 17:50:16 +0000 (10:50 -0700)]
test_checker: minor improvement in fake-server setup
This prepares for testing the differences between tubid and pubkey-based
name/longname.
david-sarah [Fri, 18 May 2012 22:56:18 +0000 (22:56 +0000)]
docs/frontends/FTP-and-SFTP.rst: remove outdated allmydata.com reference. fixes #1743
david-sarah [Wed, 16 May 2012 21:39:48 +0000 (21:39 +0000)]
Since we now require Python 2.5, we can use os.SEEK_END.
david-sarah [Wed, 16 May 2012 02:47:25 +0000 (02:47 +0000)]
Simplifications resulting from requiring Python 2.5 and therefore being able to use sqlite3 from the standard library. This also drops sqlite3 from the set of versions and paths we report.
david-sarah [Wed, 16 May 2012 02:41:49 +0000 (02:41 +0000)]
Require Python 2.5.
david-sarah [Fri, 18 May 2012 02:12:52 +0000 (02:12 +0000)]
misc/build_helpers/check-interfaces.py: avoid spurious warnings about ignored exceptions on shutdown. Also make the check function able to write errors to an arbitrary stream.
Brian Warner [Wed, 16 May 2012 23:50:57 +0000 (16:50 -0700)]
dictutil.DictOfSets: remove .union() method, it was misleading
Unlike set.union(), which returns a new set, DictOfSets.union() modified
the DictOfSets in-place. The name collision bit me when I changed some
code from using DictOfSets to a normal set, and expected that
set.union() would modify the set in-place. Since there was only one user
of DictOfSets.union, I figured it was safer to just get rid of it.
Brian Warner [Wed, 16 May 2012 23:50:43 +0000 (16:50 -0700)]
immutable repairer: populate servers-responding properly
If a server did not respond to the pre-repair filecheck, but did respond
to the repair, that server was not correctly added to the
RepairResults.data["servers-responding"] list. (This resulted from a
buggy usage of DictOfSets.union() in filenode.py).
In addition, servers to which filecheck queries were sent, but did not
respond, were incorrectly added to the servers-responding list
anyawys. (This resulted from code in the checker.py not paying attention
to the 'responded' flag).
The first bug was neatly masked by the second: it's pretty rare to have
a server suddenly start responding in the one-second window between a
filecheck and a subsequent repair, and if the server was around for the
filecheck, you'd never notice the problem. I only spotted the smelly
code while I was changing it for IServer cleanup purposes.
I added coverage to test_repairer.py for this. Trying to get that test
to fail before fixing the first bug is what led me to discover the
second bug. I also had to update test_corrupt_file_verno, since it was
incorrectly asserting that 10 servers responded, when in fact one of
them throws an error (but the second bug was causing it to be reported
anyways).
david-sarah [Wed, 16 May 2012 23:15:26 +0000 (23:15 +0000)]
Update my (davidsarah) gpg fingerprint in CREDITS. Mwahaha! :-)
david-sarah [Wed, 16 May 2012 23:08:39 +0000 (23:08 +0000)]
fileutil.py: use try/finally to close file in write_atomically.
david-sarah [Wed, 16 May 2012 22:13:29 +0000 (22:13 +0000)]
Change logging.rst to address warner's review comment. refs #1693
Brian Warner [Mon, 14 May 2012 20:32:03 +0000 (13:32 -0700)]
write node.url and portnum files atomically, to fix race in test_runner
Previously, test_runner sometimes fails because the _node_has_started()
poller fires after the portnum file has been opened, but before it has
actually been filled, allowing the test process to observe an empty file,
which flunks the test.
This adds a new fileutil.write_atomically() function (using the usual
write-to-.tmp-then-rename approach), and uses it for both node.url and
client.port . These files are written a bit before the node is really up and
running, but they're late enough for test_runner's purposes, which is to know
when it's safe to read client.port and use 'tahoe restart' (and therefore
SIGINT) to restart the node.
The current node/client code doesn't offer any better "are you really done
with startup" indicator.. the ideal approach would be to either watch the
logfile, or connect to its flogport, but both are a hassle. Changing the node
to write out a new "all done" file would be intrusive for regular
operations.
david-sarah [Mon, 14 May 2012 16:34:31 +0000 (16:34 +0000)]
Improve a comment in __init__.py.
david-sarah [Mon, 14 May 2012 03:23:52 +0000 (03:23 +0000)]
Suppress the PowmInsecureWarning from PyCrypto. refs #1586
david-sarah [Mon, 14 May 2012 00:26:37 +0000 (00:26 +0000)]
Clarify an ambiguity about which version number is meant in quickstart.rst.
Brian Warner [Sun, 13 May 2012 21:07:39 +0000 (14:07 -0700)]
performance.rst: small updates, mention (lack of) MDMF
refs #1398
Brian Warner [Sun, 13 May 2012 17:32:17 +0000 (10:32 -0700)]
CREDITS: add amiller, zooko, rearrange a bit
Brian Warner [Sun, 13 May 2012 08:12:22 +0000 (01:12 -0700)]
Doc updates and cosmetic fixes for #1115 patch.
Removes the caveat from webapi.txt about count-good-share-hosts being wrong.
This series should close #1115.
Andrew Miller [Sun, 1 Apr 2012 23:31:53 +0000 (19:31 -0400)]
Fixed an error in previous commit where an empty servermap would throw an exception in 'count-good-share-hosts'. Augmented unit test.
Signed-off-by: Andrew Miller <amiller@dappervision.com>
Andrew Miller [Sun, 1 Apr 2012 05:55:19 +0000 (01:55 -0400)]
Added tests for count-good-share-hosts under check and repair conditions. Patched the incorrect computation in immutable/filenode.py
Signed-off-by: Andrew Miller <amiller@dappervision.com>
Fixed missing import statements
Signed-off-by: Andrew Miller <amiller@dappervision.com>
Brian Warner [Sun, 13 May 2012 07:42:24 +0000 (00:42 -0700)]
test_web: fix use of headers= that's been wrong for a while
Brian Warner [Sun, 13 May 2012 07:41:53 +0000 (00:41 -0700)]
webapi: don't allow ETags in t=info or t=rename-form, both are variable
t=info contains randomly-generated ophandles, and t=rename-form contains the
name of the child being renamed, so neither is eligible for a
short-circuiting ETag. Enhanced test_web to exercise this. Had to improve
FakeCHKFileNode slightly to let it participate. Refs #443.
Brian Warner [Sun, 13 May 2012 07:19:20 +0000 (00:19 -0700)]
test_web: improve ETag tests, add If-None-Match test
Andrew Miller [Sat, 31 Mar 2012 20:46:34 +0000 (16:46 -0400)]
Added unit tests covering #466:comment-15. Refactored the 'etag' behavior for immutable files to respond to all GET '?t=' flags, not just t=None
Signed-off-by: Andrew Miller <amiller@dappervision.com>
Jeremy Fitzhardinge [Sun, 13 May 2012 03:42:52 +0000 (20:42 -0700)]
Short circuit GET on ETags match
When client does a conditional GET/HEAD with If-none-match:, if the condition
fails (ie, the client's ETag matches the file's) then we can short-circuit
the whole process and immediately return an empty body.
Jeremy Fitzhardinge [Sun, 13 May 2012 03:42:52 +0000 (20:42 -0700)]
Add ETags for immutable directories
Like immutable files, the ETag is based on the storage index. However, since
a directory is a special interpretation of a file, it is distinguished from
the file by prepending "DIR:" onto the start of the ETag, and adding
-representation on the end (where -representation is the ?t= argument, json,
info, etc).
It also checks the return of setETag and avoids generating a representation
if the client already has it.
Brian Warner [Sun, 13 May 2012 06:33:11 +0000 (23:33 -0700)]
test-dont-use-too-old-dep.py: fix tarfile timestamps
It turns out that TarFile.addfile() doesn't provide a reasonable default
timestamp, resulting in files dated to 1970 (they're probably wearing
bell-bottoms and listening to disco too). Then, when the bdist_egg command
tries to create a *zip*file with those files, it explodes because zipfiles
cannot handle timestamps before 1980 (it prefers boomboxes and jackets with
straps on the shoulders, thank you very much).
This puts a modern time.time() on the members of the tarfile, allowing future
cryptocoderarchaeologists the opportunity to make fun of fashion trends from
the user's chosen era, rather than an artificially older one.
refs #1342
david-sarah [Sat, 31 Mar 2012 22:41:22 +0000 (22:41 +0000)]
Add 'tahoe debug flogtool' command, test for --help, and docs. This version gets the help synopses more correct, and changes the doc to say that this command is added in 1.10.0 rather than 1.9.2. fixes #1693
Zooko [Sun, 13 May 2012 03:42:52 +0000 (20:42 -0700)]
modify build_helpers files
Should close #1342. This makes the actual changes to the two test
files (separated from the 'rename' patch to avoid VC complications).
Zooko [Sun, 13 May 2012 02:59:59 +0000 (19:59 -0700)]
rename build_helpers files
This is from the darcs patch for #1342, which failed to apply on my darcs
tree, so I'm landing it from git. I'm landing the rename-files part
separately from the modify-those-files part to avoid VC complications.
Brian Warner [Sun, 13 May 2012 02:19:43 +0000 (19:19 -0700)]
webapi: remove undocumented t=mkdir-p operation
Closes #380
Brian Warner [Wed, 9 May 2012 23:46:53 +0000 (16:46 -0700)]
Improve webapi t=move docs.
Brian Warner [Wed, 9 May 2012 21:18:44 +0000 (14:18 -0700)]
webui: merge 'move' form with 'rename' form
Brian Warner [Wed, 9 May 2012 21:18:37 +0000 (14:18 -0700)]
test_web: improve shouldFail2() error reporting
Brian Warner [Wed, 9 May 2012 21:18:27 +0000 (14:18 -0700)]
webapi 'move'-button cleanups
test_web.py: use shouldFail2(), safer than old shouldFail()
directory.py: forbid slashes in from_name=, return BAD_REQUEST instead of
GONE when trying to move into a non-directory
Marcus Wanner [Thu, 24 Nov 2011 06:51:42 +0000 (01:51 -0500)]
Add unit test for moving a directory
My gut tells me this case sould be tested. The rename suite tests it, so
move's will too.
Marcus Wanner [Sat, 19 Nov 2011 01:42:10 +0000 (20:42 -0500)]
Change the arbitrary URI support from implied to explicit
The move webapi function now takes a target_type argument which lets it
know whether the target is a subdirectory name or URI. This is an
improvement over the old system in which the move handler tried to guess
whether the target was a name or a URI. Also fixed a little docs
copypaste problem and tweaked some line wrapping.
Marcus Wanner [Thu, 10 Nov 2011 08:00:11 +0000 (03:00 -0500)]
Adding 'move' button to web UI, closes #1579
This adds "move file" capability to the web UI's directory display. The
support and test framework is heavily based on the similar "rename file"
feature. Unit tests and documentation are included. Multiple in-progress
versions of this patch may be found in ticket 1579. This version
includes arbitrary URI target support and is compatible with the change
from tahoe_css to tahoe.css.
Zooko O'Whielacronx [Thu, 3 May 2012 17:30:33 +0000 (17:30 +0000)]
CREDITS: lebek
david-sarah [Sun, 29 Apr 2012 02:28:44 +0000 (02:28 +0000)]
Make sure that foolscap.logging.log.setLogDir is called with a str (not unicode) path, v2. Includes test. fixes #1725
Brian Warner [Tue, 24 Apr 2012 05:37:28 +0000 (01:37 -0400)]
introweb announcements: show serverid, not tubid
'serverid' is the pubkey (for V2 clients), falling back to the tubid (for V1
clients). This also required cleaning up the way the index is created for the
old V1 introducer.
Brian Warner [Mon, 23 Apr 2012 22:02:22 +0000 (18:02 -0400)]
Fix introweb display for mixed V1/V2 clients. Closes #1721.
This significantly cleans up the IntroducerServer web-status renderers.
Instead of poking around in the introducer's internals, now the web-status
renderers get clean AnnouncementDescriptor and SubscriberDescriptor
objects. They are still somewhat foolscap-centric, but will provide a clean
abstraction boundary for future improvements.
The specific #1721 bug was that old (V1) subscribers were handled by
wrapping their RemoteReference in a special WrapV1SubscriberInV2Interface
object, but the web-status display was trying to peek inside the object to
learn what host+port it was associated with, and the wrapper did not proxy
those extra attributes.
A test was added to test_introducer to make sure the introweb page renders
properly and at least contains the nicknames of both the V1 and V2 clients.
Patrick R McDonald [Wed, 11 Apr 2012 02:36:43 +0000 (21:36 -0500)]
Updated webapi.rst to list /cap as a synonym for /uri
markus reichelt [Sat, 14 Apr 2012 14:01:07 +0000 (16:01 +0200)]
Adding jg71 to CREDITS
Patrick R McDonald [Tue, 10 Apr 2012 14:55:25 +0000 (09:55 -0500)]
Fix for ticket #1662
david-sarah [Fri, 6 Apr 2012 04:39:46 +0000 (04:39 +0000)]
docs/frontends/drop-upload.rst: document more known issues and link to new ticket for an existing one.
david-sarah [Thu, 5 Apr 2012 23:57:23 +0000 (23:57 +0000)]
Change capitalization of WUI and introducer welcome page headings; add test for introducer welcome page. Also fix a typo in a CSS class name. fixes #1708
Brian Warner [Wed, 4 Apr 2012 19:11:03 +0000 (12:11 -0700)]
test/common.py: remove ununsed 'is_bad' mechanism
This was a premature feature addition to the mock filenode, and gets in the
way of the IServer refactoring I'm trying to do. Best to remove it now and
re-introduce it in a better form later when it's actually needed.
Brian Warner [Wed, 4 Apr 2012 19:05:31 +0000 (12:05 -0700)]
checker.py: minor simplifications
Brian Warner [Wed, 4 Apr 2012 18:14:09 +0000 (11:14 -0700)]
make IServer instances retain identity in copy() and deepcopy()
Brian Warner [Wed, 4 Apr 2012 18:13:59 +0000 (11:13 -0700)]
move IServer from storage_client.py to interfaces.py
Brian Warner [Tue, 3 Apr 2012 03:02:59 +0000 (20:02 -0700)]
Rename web CheckResults to -Renderer, to avoid confusion. Closes #1705.
This avoids the name collision between the actual results
objects (defined in allmydata.check_results) and the code that renders
these objects into HTML (defined in allmydata.web.check_results). Only
the web-side objects were renamed.
Brian Warner [Mon, 2 Apr 2012 23:32:05 +0000 (16:32 -0700)]
webapi.rst: de-tabify
Brian Warner [Sun, 1 Apr 2012 22:08:42 +0000 (15:08 -0700)]
servermap.py: oops, fix _done() condition, good catch by davidsarah
Zooko O'Whielacronx [Sun, 1 Apr 2012 21:40:39 +0000 (21:40 +0000)]
doc: cross-link known_issues.rst and cautions.rst with one another
Zooko O'Whielacronx [Sun, 1 Apr 2012 21:20:02 +0000 (21:20 +0000)]
docs: FTP-and-SFTP.rst: recommend SFTP
Add an explicit recommendation of SFTP over FTP. Separate the known issues of
FTP from SFTP. List "SFTP" first in all lists of the two. Use unicode bullet
points and prepend a utf-8 BOM. Use out-of-line rst hyperlinks.
david-sarah [Sun, 14 Aug 2011 22:59:59 +0000 (22:59 +0000)]
interfaces.py: ensure that NoSuchChildError can be converted to str even when it is for a non-ASCII name. fixes #1483
david-sarah [Sun, 1 Apr 2012 02:28:26 +0000 (02:28 +0000)]
misc/build_helpers/show-tool-versions.py: s/print_stderr/print_stdout/
david-sarah [Wed, 17 Aug 2011 22:26:51 +0000 (22:26 +0000)]
bin/tahoe-script.template: fix the error message that is displayed when a runner script cannot be found. fixes #1488
Zooko O'Whielacronx [Sun, 1 Apr 2012 01:57:17 +0000 (01:57 +0000)]
docs: quickstart: edits
• use out-of-line links to avoid a warning from rst2html --verbose (fixes #1704)
• reflow to 77 fill-column and prepend utf-8 BOM (fixes #1703)
• recommend Python 2.7 (fixes #1702)
• remove link to wiki:AdvancedInstall (fixes #1701)
david-sarah [Sun, 1 Apr 2012 01:36:55 +0000 (01:36 +0000)]
Spelling error in a comment.
david-sarah [Mon, 8 Aug 2011 18:05:52 +0000 (18:05 +0000)]
test_node.py: test that we tolerate a UTF-8 BOM at the start of tahoe.cfg, and can read UTF-8 option values. refs #1470
david-sarah [Mon, 8 Aug 2011 18:02:04 +0000 (18:02 +0000)]
node.py: tolerate a UTF-8 BOM at the start of tahoe.cfg. fixes #1470
david-sarah [Wed, 14 Sep 2011 14:39:47 +0000 (14:39 +0000)]
mutable/layout.py: improve confusing documentation of layout. fixes #1534
Zooko O'Whielacronx [Sun, 1 Apr 2012 00:59:25 +0000 (00:59 +0000)]
setup: show-tool-versions: report cl only on windows, report buildslave, git, openssl, and lzip, but not 7za
david-sarah [Sat, 31 Mar 2012 22:39:34 +0000 (22:39 +0000)]
Document PYTHONPATH problem when running flogtool. refs #1693
Brian Warner [Sat, 31 Mar 2012 06:56:41 +0000 (23:56 -0700)]
Mutable repair: use new MODE_REPAIR to query all servers *and* get privkey
This fixes bug #1689. Repair was using MODE_READ to build the servermap,
which doesn't try hard enough to grab the privkey, and also doesn't guarantee
sending queries to all servers. This patch adds a new MODE_REPAIR which does
both, and does a separate, distinct mapupdate to start wth repair cycle,
instead of relying upon the (MODE_CHECK) mapupdate leftover from the
filecheck that triggered the repair.
Brian Warner [Sat, 31 Mar 2012 18:36:20 +0000 (11:36 -0700)]
Add test for bug #1689: repairing empty file hits no-privkey assertion
david-sarah [Sat, 31 Mar 2012 02:32:47 +0000 (02:32 +0000)]
FTP-and-SFTP.rst: there were two more instances of 'rootcap'. Also made the wording tweak from ticket:1487#comment:4 . fixes #1487
david-sarah [Sat, 31 Mar 2012 02:17:25 +0000 (02:17 +0000)]
test_ftp.py: fix a couple of unused imports. refs #1668
david-sarah [Sat, 31 Mar 2012 01:37:30 +0000 (01:37 +0000)]
FTP-and-SFTP.rst: directories containing mutable files should now be listable via FTP. refs #680
Peter Le Bek [Thu, 22 Mar 2012 13:18:06 +0000 (13:18 +0000)]
ftpd file `size' attribute must be an integer
Peter Le Bek [Fri, 30 Mar 2012 23:41:39 +0000 (23:41 +0000)]
unit test for ftpd LIST
Peter Le Bek [Fri, 30 Mar 2012 23:41:19 +0000 (23:41 +0000)]
fix ftpd mtime retrieval
Brian Warner [Sat, 31 Mar 2012 00:29:06 +0000 (17:29 -0700)]
test_introducer.SystemTest: fix race condition
SystemTest has a couple of different phases, separated by a poller which
waits for everything to be idle (all messages delivered, none in flight). It
does this by watching some internal "_debug_outstanding" counters in the
server and in each client, and waiting for them to hit zero.
Just before the last phase, we replace the server with a new one (to make
sure clients re-send their messages properly). Unfortunately, the polling
function closed over the variable holding the original server, and didn't see
the replacement. It kept polling the old server, and failed to notice the
outstanding messages for the new server. The last phase of the test (check3)
was started too early, which failed (since some messages had not yet been
delivered), and then exploded in a flurry of dirty-reactor errors (because
some messages were delivered after test shutdown).
This replaces the closed-over-variable with a "self.the_introducer", which
seems to fix the race.
One additional place to look at in the future: the client
announcement-receive path (remote_announce) uses an eventually(). If the
message has been received and the eventual-send posted (but not yet executed)
when the poller sees it, the poller might erroneously conclude that the
client is idle and cause the same problem as above. To fix this, the poller
(probably all pollers) could be enhanced to do a flushEventualQueue before
querying the are-we-done-yet predicate function.
david-sarah [Thu, 22 Mar 2012 22:05:34 +0000 (22:05 +0000)]
Cosmetic formatting in docs.
david-sarah [Thu, 22 Mar 2012 22:04:53 +0000 (22:04 +0000)]
Put SFTP before FTP in various docs. fixes #1692
david-sarah [Thu, 22 Mar 2012 22:01:18 +0000 (22:01 +0000)]
Correct a link to frontends/drop-upload.rst. fixes #1690
Brian Warner [Sat, 17 Mar 2012 23:52:32 +0000 (16:52 -0700)]
Fix mutable status (mapupdate/retrieve/publish) to use serverids, not tubids
This still leaves immutable-publish results incorrectly using tubids instead
of serverids. That will need some more work, since it might change the Helper
interface.
Brian Warner [Wed, 14 Mar 2012 23:36:40 +0000 (16:36 -0700)]
IServer.get_name(): remove v0- prefix from displayed server names
Don't remove the prefix if it isn't there: that avoids the need to fix tests
which use a bogus key (usually all-zeros).
david-sarah [Wed, 14 Mar 2012 23:50:40 +0000 (23:50 +0000)]
Fix a missing comma in the last patch. refs #1295
david-sarah [Wed, 14 Mar 2012 23:47:29 +0000 (23:47 +0000)]
Temporarily suppress the DeprecationWarning about IFinishableConsumer; it's irritating, but not in a way that is likely to make me fix the underlying issue (#1525) any sooner :-). refs #1295
Brian Warner [Wed, 14 Mar 2012 06:13:51 +0000 (23:13 -0700)]
minor: hush pyflakes, move pycryptopp dep to unconditional section
Also change Makefile's "pyflakes" rule to emit less output, so buildbot will
count errors properly.
david-sarah [Tue, 13 Mar 2012 20:30:41 +0000 (20:30 +0000)]
Update find_links URLs in setup.cfg to https://tahoe-lafs.org. This is not just a doc change; look out for compatibility problems.
david-sarah [Tue, 13 Mar 2012 20:28:53 +0000 (20:28 +0000)]
Minor updates to URLs.