Alberto Berti
89a20fe4dd
Better implementation of filtering algorithm.
2009-02-22 16:40:49 -07:00
Alberto Berti
2fbca1596b
Removed '.hgrags' from vcs excludes
2009-02-22 16:39:46 -07:00
Alberto Berti
5d629d0050
Added tests for the --exclude* options of backup command.
2009-02-22 10:51:06 -07:00
Alberto Berti
5d57da93fd
Added --exclude, --exclude-from and --exclude-vcs options to backup command.
...
It is still impossible to specify absolute exclusion path, only
relative. I must check with tar or rsync how they allow them to be
specified.
2009-02-22 11:08:29 -07:00
Alberto Berti
c54fdd61b0
Raise a more explanatory exception for errors encountered during backup processing.
2009-02-22 11:02:52 -07:00
Alberto Berti
ff6a77cc6b
Added tests for the fixed alias related command's synopsis
2009-02-22 10:37:32 -07:00
Alberto Berti
ffab4002aa
Add missing synopsis and descriptions for alias commands.
2009-02-20 18:31:06 -07:00
Brian Warner
c85eccbf64
test_web.py: increase test coverage of web.status.plural()
2009-02-21 18:01:16 -07:00
Brian Warner
77f3b83d68
crawler: fix performance problems: only save state once per timeslice (not after every bucket), don't start the crawler until 5 minutes after node startup
2009-02-21 14:56:49 -07:00
Brian Warner
a04d3b8fe8
test_system: oops, don't assume that all files in storage/ are in a deep storage/shares/prefix/si/shnum path, since now the crawler pickle has a short path
2009-02-21 00:17:10 -07:00
Brian Warner
b9c4f4bdf6
crawler: tolerate low-resolution system clocks (i.e. windows)
2009-02-21 00:15:33 -07:00
Brian Warner
106d31b112
BucketCountingCrawler: store just the count, not cycle+count, since it's too easy to make usage mistakes otherwise
2009-02-20 21:58:31 -07:00
Brian Warner
3c6471c717
test_storage: startService the server, as is now the standard practice
2009-02-20 21:57:55 -07:00
Brian Warner
f934289d2d
crawler: load state from the pickle in init, rather than waiting until startService, so get_state() can be called early
2009-02-20 21:57:20 -07:00
Brian Warner
1077826357
BucketCountingCrawler: rename status and state keys to use 'bucket' instead of 'share', because the former is more accurate
2009-02-20 21:46:06 -07:00
Brian Warner
d2d297f12f
storage: also report space-free-for-root and space-free-for-nonroot, since that helps users understand the space-left-for-tahoe number better
2009-02-20 21:28:56 -07:00
Brian Warner
b3cd4952bd
storage: add bucket-counting share crawler, add its output (number of files+directories maintained by a storage server) and status to the webapi /storage page
2009-02-20 21:04:08 -07:00
Brian Warner
d14f00c537
storage: move si_b2a/si_a2b/storage_index_to_dir out of server.py and into common.py
2009-02-20 21:03:09 -07:00
Brian Warner
73e05bf967
crawler: add get_progress, clean up get_state
2009-02-20 18:27:43 -07:00
Brian Warner
2e45619844
web/storage: make sure we can handle platforms without os.statvfs too
2009-02-20 16:03:53 -07:00
Brian Warner
c6a061e600
crawler: provide for one-shot crawlers, which stop after their first full cycle, for share-upgraders and database-populaters
2009-02-20 15:19:11 -07:00
Brian Warner
00677ff9a5
web: add Storage status page, improve tests
2009-02-20 14:29:26 -07:00
Brian Warner
ff6907a557
storage: include reserved_space in stats
2009-02-20 14:29:20 -07:00
Brian Warner
4b8cc42f18
web/check_results: sort share identifiers in the sharemap display
2009-02-20 12:29:22 -07:00
Brian Warner
c6c30ac5d4
webapi: pass client through constructor arguments, remove IClient, should make it easier to test web renderers in isolation
2009-02-20 12:15:54 -07:00
Brian Warner
e357b8a679
test/no_network: do startService on the storage servers, make it easier to customize the storage servers
2009-02-19 20:22:54 -07:00
Brian Warner
ef4ff21ae7
crawler: modify API to support upcoming bucket-counting crawler
2009-02-19 19:31:42 -07:00
Brian Warner
9bc08158c6
test_backupdb: make the not-a-database file larger, since the older sqlite-2.3.2 on OS-X is easily fooled
2009-02-19 18:04:09 -07:00
Brian Warner
26fbe9df95
web/reliability: add parameter descriptions, adapted from a patch from Terrell Russell.
2009-02-19 16:29:18 -07:00
Brian Warner
2f751332e8
test_crawler: hush pyflakes
2009-02-19 14:23:40 -07:00
Brian Warner
ff9811211d
test_crawler: disable the percentage-of-cpu-used test, since it is too unreliable on our slow buildslaves. But leave the code in place for developers to run by hand.
2009-02-19 14:16:54 -07:00
Brian Warner
790a10d1b2
reliability.py: fix the numpy conversion, it was completely broken. Thanks to Terrell Russell for the help.
2009-02-19 13:55:15 -07:00
Brian Warner
268cd538e2
reliability: switch to NumPy, since Numeric is deprecated
2009-02-19 01:44:35 -07:00
Brian Warner
f10aa59e5b
test_crawler: don't require >=1 cycle on cygwin
2009-02-19 00:58:18 -07:00
Brian Warner
dfd72c6483
crawler: use fileutil.move_info_place in preference to our own version
2009-02-18 23:13:42 -07:00
Brian Warner
89f041ac83
fileutil: add move_into_place(), to perform the standard unix trick of atomically replacing a file, with a fallback for windows
2009-02-18 23:13:10 -07:00
Brian Warner
b949ea4f32
crawler: fix problems on windows and our slow cygwin slave
2009-02-18 22:24:31 -07:00
Brian Warner
193889f793
#633 : first version of a rate-limited interruptable share-crawler
2009-02-18 21:46:33 -07:00
Brian Warner
4646451de6
change StorageServer to take nodeid in the constructor, instead of assigning it later, since it's cleaner and because the original problem (Tubs not being ready until later) went away
2009-02-18 16:23:01 -07:00
Brian Warner
b398a69f14
test_system: split off checker tests to test_deepcheck.py, this file is too big
2009-02-18 15:42:34 -07:00
Brian Warner
ef53da2b12
break storage.py into smaller pieces in storage/*.py . No behavioral changes.
2009-02-18 14:46:55 -07:00
Brian Warner
a0c5f92cbd
immutable/layout: minor change to repr name
2009-02-18 14:46:48 -07:00
Brian Warner
2346d8621d
interfaces.py: allow add/renew/cancel-lease to return Any, so that 1.3.1 clients (the first to use these calls) can tolerate future storage servers which might return something other than None
2009-02-18 13:29:03 -07:00
Brian Warner
bce4a5385b
add --add-lease to 'tahoe check', 'tahoe deep-check', and webapi.
2009-02-17 19:32:43 -07:00
Brian Warner
e9563ebc02
change RIStorageServer.remote_add_lease to exit silently in case of no-such-bucket, instead of raising IndexError, because that makes the upcoming --add-lease feature faster and less noisy
2009-02-17 19:30:53 -07:00
Brian Warner
fde2289e7b
CLI #590 : convert 'tahoe deep-check' to streaming form, improve display, add tests
2009-02-17 17:15:11 -07:00
Brian Warner
0e78b2587c
interfaces.py: document behavior of add_lease/renew_lease/cancel_lease, before I change it
2009-02-17 13:48:09 -07:00
Brian Warner
d52eadfc58
test_backupdb: improve error messages if the test fails
2009-02-17 11:08:38 -07:00
Brian Warner
476a5c8fac
webapi #590 : add streaming deep-check. Still need a CLI tool to use it.
2009-02-16 23:35:53 -07:00
Brian Warner
8579e25059
test_web.Grid: change the CHECK() function to make it easier to test t= values with hyphens in them
2009-02-16 23:00:34 -07:00
Brian Warner
522641bb32
test_web: improve checker-results coverage with a no-network -based test, enhance no-network harness to assist, fix some bugs in web/check_results.py that were exposed
2009-02-16 22:12:42 -07:00
Brian Warner
059f24adca
web: fix handling of reliability page when Numeric is not available
2009-02-16 19:56:58 -07:00
Brian Warner
f2f8ede11f
test/no_network: update comments with setup timing: no_network takes 50ms, SystemTestMixin takes 2s (on my laptop)
2009-02-16 18:06:43 -07:00
Brian Warner
040cb39613
test_upload: rewrite in terms of no-network GridTestMixin, improve no_network.py as necessary
2009-02-16 17:44:57 -07:00
Brian Warner
911abcc34b
test_download: rewrite in terms of no-network GridTestMixin, improve no_network.py as necessary
2009-02-16 17:36:58 -07:00
Brian Warner
e9427ea777
test_dirnode.py: convert Deleter to new no-network gridtest
2009-02-16 17:23:48 -07:00
Brian Warner
267d91f797
test_cli.py: modify to use the new 'no-network' gridtest instead of SystemTestMixin, which speeds it up from 73s to 43s on my system
2009-02-16 17:20:05 -07:00
Brian Warner
191fd8556a
tests: fix no_network framework to work with upload/download and checker
2009-02-16 17:19:47 -07:00
Brian Warner
0409bc5628
client.py: improve docstring
2009-02-16 17:15:32 -07:00
Brian Warner
770f976a85
test_cli: add test coverage for help strings
2009-02-16 15:08:33 -07:00
Brian Warner
ebe0f2ac08
test/no_network: new test harness, like system-test but doesn't use the network so it's faster
2009-02-16 14:58:44 -07:00
Brian Warner
8ff76c6269
interfaces.py: minor docstring edit
2009-02-16 14:58:16 -07:00
Brian Warner
df90dd8e73
test_cli: improve test coverage slightly
2009-02-15 21:04:51 -07:00
Brian Warner
3235b9630b
test_util: get almost full test coverage of dictutil, starting with the original pyutil tests as a base. The remaining three uncovered lines involve funny cases of ValueOrderedDict that I can't figure out how to get at
2009-02-15 20:32:10 -07:00
Brian Warner
cc27b218f5
provisioning/reliability: add tests, hush pyflakes, remove dead code, fix web links
2009-02-15 16:24:51 -07:00
Brian Warner
e1380b132b
util/statistics: add tests, fix mean_repair_cost
2009-02-15 16:23:26 -07:00
Brian Warner
812f23585d
test_repairer: hush pyflakes
2009-02-15 16:23:10 -07:00
Brian Warner
e2efd911a3
build a 'reliability' web page, with a simulation of file decay and repair over time
2009-02-13 17:42:34 -07:00
Shawn Willden
51ab76875b
More lossmodel work, on repair.
2009-01-15 20:56:48 -07:00
Shawn Willden
21832280da
Statistics module
...
Added a statistics module for calculating various facets of
share survival statistics.
2009-01-13 20:12:35 -07:00
Brian Warner
60b06b8d32
test_dirnode: #625 run deep-check on a readonly dirnode too
2009-02-13 14:53:37 -07:00
Brian Warner
4aee51ff6f
figleaf_htmlizer: fix order of summary counters
2009-02-13 09:57:53 -07:00
Brian Warner
aee319e236
figleaf_htmlizer: oops, re-ignore files that aren't under root, like code in auto-built eggs
2009-02-13 00:00:22 -07:00
Zooko O'Whielacronx
14d23d99a5
setup: make the "full version string" be "allmydata-tahoe/1.3.0" instead of "allmydata-tahoe-1.3.0" and the UserAgent string of the cli be "allmydata-tahoe/1.3.0 (tahoe-client)"
...
This is webbish. Thanks to kpreid for suggesting it.
2009-02-12 22:37:38 -07:00
Brian Warner
30724593fc
figleaf_htmlizer: emit stats to stdout, so buildbot can see it
2009-02-12 15:10:20 -07:00
Brian Warner
781c271b82
figleaf_htmlizer: render changes in coverage relative to a previous test run using --old-coverage
2009-02-12 15:04:12 -07:00
Zooko O'Whielacronx
d7dbd6675e
immutable repairer: fix DownUpConnector so that it satisfies short reads the were requested after the last write and before the close
...
This is probably the cause of the very rare "loss of progress" bug. This is tested by unit tests. A recent patch changed this to errback instead of losing progress, and now this patch is changing it again to return a short read instead of errbacking. Returning a short read is what the uploader (in encode.py) is expecting, when it is reading the last block of the ciphertext, which might be shorter than the other blocks.
2009-02-12 17:04:47 -07:00
Zooko O'Whielacronx
bdb992467c
immutable repairer: add an assertion that a certain value in this tricky function is always what I think it is
2009-02-12 16:31:32 -07:00
Zooko O'Whielacronx
2d45d192d1
immutable repairer: add tests of how it handles short reads
2009-02-12 16:48:53 -07:00
Zooko O'Whielacronx
8303f90da0
nodeadmin: node stops itself if a hotline file hasn't been touched in 60 seconds now, instead of in 40 seconds
...
A test failed on draco (MacPPC) because it took 49 seconds to get around to running the test, and the node had already stopped itself when the hotline file was 40 seconds old.
2009-02-12 15:21:31 -07:00
Zooko O'Whielacronx
76d7cc4404
immutable repairer: errback any pending readers of DownUpConnectorwhen it runs out of bytes, and test that fact
2009-02-11 20:11:29 -07:00
Brian Warner
125bf09528
figleaf_htmlizer: more rearranging, behavior should still be unchanged
2009-02-11 20:05:15 -07:00
Brian Warner
e02ecec1ad
figleaf_htmlizer: break it up into subfunctions, behavior should still be unchanged
2009-02-11 19:56:07 -07:00
Brian Warner
6bde1b0345
figleaf_htmlizer: rewrite in class form, behavior should be the same as before
2009-02-11 19:40:50 -07:00
Brian Warner
6db747b1f0
figleaf_htmlizer: rewrite with twisted.python.usage, remove logging: should behave the same as before
2009-02-11 19:16:43 -07:00
Brian Warner
f3ed579e74
figleaf_htmlizer: expand tabs, fix to 4-space indents. No functional changes.
2009-02-11 19:05:42 -07:00
Zooko O'Whielacronx
7eb260a9cf
versioning: include an "appname" in the application version string in the versioning protocol, and make that appname be controlled by setup.py
...
It is currently hardcoded in setup.py to be 'allmydata-tahoe'. Ticket #556 is to make it configurable by a runtime command-line argument to setup.py: "--appname=foo", but I suddenly wondered if we really wanted that and at the same time realized that we don't need that for tahoe-1.3.0 release, so this patch just hardcodes it in setup.py.
setup.py inspects a file named 'src/allmydata/_appname.py' and assert that it contains the string "__appname__ = 'allmydata-tahoe'", and creates it if it isn't already present. src/allmydata/__init__.py import _appname and reads __appname__ from it. The rest of the Python code imports allmydata and inspects "allmydata.__appname__", although actually every use it uses "allmydata.__full_version__" instead, where "allmydata.__full_version__" is created in src/allmydata/__init__.py to be:
__full_version__ = __appname + '-' + str(__version__).
All the code that emits an "application version string" when describing what version of a protocol it supports (introducer server, storage server, upload helper), or when describing itself in general (introducer client), usese allmydata.__full_version__.
This fixes ticket #556 at least well enough for tahoe-1.3.0 release.
2009-02-11 17:18:16 -07:00
Brian Warner
23fdd4f48e
backupdb: cosmetic: capitalize the no-pysqlite instructions properly. Thanks to Terrell Russell for the catch.
2009-02-11 15:28:30 -07:00
Brian Warner
5f288842d5
test_repairer: disable repair-from-corruption tests until other things are improved well enough to make it useful
2009-02-11 15:01:59 -07:00
Zooko O'Whielacronx
079c1851a9
trivial: whitespace
2009-02-11 11:19:35 -07:00
Brian Warner
7d907f35aa
storage: make add-lease work, change default ownernum=1 since 0 is reserved to mean 'no lease here'
2009-02-10 23:39:38 -07:00
Brian Warner
13a3ef5ec1
#620 : storage: allow mutable shares to be deleted, with a writev where new_length=0
2009-02-10 23:37:56 -07:00
Brian Warner
6a5f28f47d
test_cli.Backup: insert some stalls to make sure two successive backups get distinct timestamps, avoiding intermittent failures
2009-02-10 20:37:09 -07:00
Brian Warner
f7a263eb0b
#619 : make 'tahoe backup' complain and refuse to run if sqlite is unavailable and --no-backupdb is not passed
2009-02-10 18:49:10 -07:00
Brian Warner
3b1c8ceb86
tests: increase the default timeout for SystemTestMixin -based tests to 300 seconds, since our slower buildslaves sometimes take longer than the default 120s
2009-02-10 14:44:12 -07:00
Zooko O'Whielacronx
ef1bfdd2bf
immutable: repairer: add a simple test to exercise the "leftover" code path, fix the bug (and rename the variable "leftover" to "extra")
2009-02-10 12:12:45 -07:00
Zooko O'Whielacronx
75e4e67ed7
immutable: tighten preconditions -- you can write empty strings or read zero bytes, and add the first simple unit test of DownUpConnector
2009-02-10 00:56:47 -07:00
Zooko O'Whielacronx
d436c6235d
immutable: tests: the real WRITE_LEEWAY is 35 (it was a mistake to move it from 10 to 35 earlier -- I had seen a failure in which it took 35 times as many writes as I thought were optimal, but I misread and thought it took only 20 times as many)
2009-02-09 23:53:48 -07:00
Zooko O'Whielacronx
c59940852b
immutable: defensive programming: assert that the encrypted readable gave you no more than the number of bytes you asked for
...
(There is a bug in the current DownUpConnector which can cause it to give more bytes than you asked for on one request, and then less on the next, effectively shifting some of the bytes to an earlier request, but I think this bug never gets triggered in practice.)
2009-02-09 23:46:05 -07:00
Zooko O'Whielacronx
65c12b24b8
immutable: tests: sigh, raise, again the limit of how many extra writes you can do and still pass this test
...
Obviously requiring the code under test to perform within some limit isn't very meaningful if we raise the limit whenever the test goes outside of it.
But I still don't want to remove the test code which measures how many writes (and, elsewhere, how many reads) a client does in order to fulfill these duties.
Let this number -- now 20 -- stand as an approximation of the inefficiency of our code divided by my mental model of how many operations are actually optimal for these duties.
2009-02-09 20:09:31 -07:00
