Commit Graph

1106 Commits

Author SHA1 Message Date
Brian Warner
f4c048bbeb encode.py: clean up handling of lost peers during upload, add some logging 2007-06-06 12:40:16 -07:00
Brian Warner
daa4c32381 runner.py: allow all directory-using commands (create/start/stop) to accept argv[-1] as well as --basedir 2007-06-06 11:37:19 -07:00
Brian Warner
6bb9debc16 encode: tolerate lost peers, as long as we still get enough shares out. Closes #17. 2007-06-06 10:32:40 -07:00
Brian Warner
3dfd26970b move validation data to thingA, URI has storage_index plus thingA hash
This (compatibility-breaking) change moves much of the validation data and
encoding parameters out of the URI and into the so-called "thingA" block
(which will get a better name as soon as we find one we're comfortable with).
The URI retains the "storage_index" (a generalized term for the role that
we're currently using the verifierid for, the unique index for each file
that gets used by storage servers to decide which shares to return), the
decryption key, the needed_shares/total_shares counts (since they affect
peer selection), and the hash of the thingA block.

This shortens the URI and lets us add more kinds of validation data without
growing the URI (like plaintext merkle trees, to enable strong incremental
plaintext validation), at the cost of maybe 150 bytes of alacrity. Each
storage server holds an identical copy of the thingA block.

This is an incompatible change: new messages have been added to the storage
server interface, and the URI format has changed drastically.
2007-06-01 18:48:01 -07:00
Brian Warner
d1d7cdd859 node.py: multi-class exception calls need parentheses 2007-05-31 18:32:21 -07:00
Brian Warner
7124f94461 download.py: refactor bucket_failed() a bit, add some docs 2007-05-31 18:31:36 -07:00
Brian Warner
edaed9c32c test_system.py: minor reformatting 2007-05-31 18:31:01 -07:00
Brian Warner
4f0e944087 node.py: add logging of startup/shutdown, for the cygwin test_system failure 2007-05-31 13:44:22 -07:00
Brian Warner
8875adf807 test_system.py: add a log message to help track down the occasional cygwin failure 2007-05-31 12:01:14 -07:00
Brian Warner
1160c1def2 node.py: log twisted version along with tahoe/foolscap/zfec versions 2007-05-31 11:21:06 -07:00
Brian Warner
ea78b4b605 check_memory: getting closer, now we have memusage numbers for uploads of 10kB and 10MB files 2007-05-29 17:39:39 -07:00
Zooko O'Whielacronx
04b649f971 switch from FieldStorage.value to FieldStorage.file
Unfortunately this doesn't make the O(n) memory usage go away.  It might reduce the constants -- I'm not sure.  I look forward to enhancement #54 -- memory usage tests!
2007-05-25 16:00:19 -07:00
Zooko O'Whielacronx
763cfc7874 amdlib.util: merge in changes to humanreadable.py that were made in pyutil 2007-05-25 15:49:57 -07:00
Brian Warner
c405c6117f check_memory.py: finish the failsafe-shutdown code 2007-05-24 17:34:42 -07:00
Brian Warner
4c7c5df41e runner.py: spoke too soon. Really fix #51 this time. 2007-05-24 11:20:39 -07:00
Brian Warner
d0db98cc40 runner.py: expanduser() basedirs, so '~' works. Closes #51. 2007-05-24 11:10:19 -07:00
Brian Warner
8fc1bf4330 test_node.py: hush pyflakes warnings 2007-05-23 17:55:04 -07:00
Brian Warner
cbca430062 node.py: hush pyflakes warnings 2007-05-23 17:54:48 -07:00
Brian Warner
2d21029367 test_introducer_and_vdrive: remove the assumption that startService returns a Deferred 2007-05-23 17:37:20 -07:00
Brian Warner
83daa265ba test_system.py: check size in web-ui display. closes #37 2007-05-23 17:25:49 -07:00
Zooko O'Whielacronx
80b09b6d54 add in-line doc that Josh wrote as he was trying to understand this code 2007-05-23 15:11:23 -07:00
Zooko O'Whielacronx
44902c5152 add unit test for "advertised_ip_addresses" feature and fix bug in that feature uncovered by this unit test 2007-05-23 15:08:55 -07:00
Zooko O'Whielacronx
77f55f368a make stopService() defer until startService() completes (fixes a problem with the new not-yet-committed unit test) 2007-05-23 15:08:03 -07:00
Brian Warner
f9e45391f6 node.py: use 'node.pem' for all nodes
Rather than use separate client.pem and introducer.pem files, use 'node.pem'
for all nodes regardless of what type it is. This is slightly cleaner, but
introduces a compatibility. Users who upgrade to this change should do
'mv client.pem node.pem' to avoid generating a new certificate and thus
changing their TubID.
2007-05-23 12:48:52 -07:00
Brian Warner
9edc6574ae node.py: use Tub's certFile= argument instead of doing it ourselves 2007-05-23 12:41:23 -07:00
Zooko O'Whielacronx
3872e94da2 rename all "*PBURL*" to "*FURL*"
This breaks backwards compatibility with Tahoe v0.2 -- the first public release of Tahoe.
2007-05-22 14:08:30 -07:00
Zooko O'Whielacronx
9cd1757799 refactor iputil and make it return addresses in descending order of goodness instead of in a set
Actually of course iputil can't tell exactly how good they are, and a wise user
of iputil will try all of them.  But you can't try all of them simultaneously,
so you might as well try the best ones first.
2007-05-22 14:06:37 -07:00
Zooko O'Whielacronx
921e643b93 don't check for existence of portnum file and then try to open it -- instead try to open it and catch exception
This avoids a race condition, also known as time-of-check-to-time-of-use.
2007-05-22 14:06:00 -07:00
Zooko O'Whielacronx
98066ccaee don't test for existence of certfile and then try to open it -- instead try to open it and catch exception
This avoids a race condition, also known as time-of-check-to-time-of-use.
2007-05-22 14:04:16 -07:00
Zooko O'Whielacronx
a2970cf7c1 fix handling of local_ip file and rename it to advertised_ip_addresses and document it in README 2007-05-22 14:01:40 -07:00
Zooko O'Whielacronx
10df6fac02 raise constraint on FURLs from 150 chars to 1000 chars 2007-05-22 13:59:17 -07:00
Brian Warner
05163ec8e1 change uri-packer-unpacker to deal with dictionaries, not fragile tuples 2007-05-23 11:18:49 -07:00
Zooko O'Whielacronx
3d1adf6d30 update docs, remove extraneous licence text, sort module names in import statement
closes #46 ?
2007-05-21 13:42:51 -07:00
Zooko O'Whielacronx
9b70d17d4b enable private upload, in which the file is inserted at the grid layer but not at the vdrive layer
This patch is actually by Faried Nawaz, as per ticket #33:

http://allmydata.org/trac/tahoe/ticket/33
2007-05-16 08:40:19 -07:00
Brian Warner
82ba0f8540 UNDO: forget about old peers (closes #26)
Add a new method to RIIntroducer, to allow the central introducer node to
remove peers from the active set after they've gone away. Without this,
client nodes accumulate stale peer FURLs forever. This introduces a
compatibility break, as old introducers won't know about the 'lost_peers'
message, although the errors produced are probably harmless.
2007-05-07 19:10:24 -07:00
Brian Warner
04af4a48b5 forget about old peers (closes #26)
Add a new method to RIIntroducer, to allow the central introducer node to
remove peers from the active set after they've gone away. Without this,
client nodes accumulate stale peer FURLs forever. This introduces a
compatibility break, as old introducers won't know about the 'lost_peers'
message, although the errors produced are probably harmless.
2007-05-07 19:10:24 -07:00
Zooko O'Whielacronx
e0bedb64ea manually merge recent changes to pyutil's copy of "version.py" with allmydata's copy of "version.py" 2007-05-09 16:04:05 -07:00
Faried Nawaz
2016b846b9 display file size in directory.xhtml 2007-05-04 13:07:32 -07:00
Brian Warner
5b3a1cdc56 test_introducer.py: turn off the defer.setDebugging(True) call that was left over from a troubleshooting session: it slows down the test suite by a factor of 2 2007-05-04 00:45:23 -07:00
Brian Warner
9bd213abe7 version.py: handle both '0.2.0' and '0.2.0-1', remove 'tags' for now 2007-05-03 20:31:27 -07:00
Brian Warner
3f6b660272 update version-number handling, pull release tags from darcs history 2007-05-03 20:14:07 -07:00
Zooko O'Whielacronx
5610a1b2ce tahoe v0.2.0-0-UNSTABLE 2007-05-02 16:23:23 -07:00
Zooko O'Whielacronx
ff08abb67e bump test timeout even higher
It does indeed take longer than 2400 seconds to run test_upload_and_download on a virtual windows machine when the underlying real machine is heavily loaded down with filesystem analysis runs...
2007-05-02 15:31:05 -07:00
Brian Warner
0650e5543f figleaf_htmlizer: when all counts are zero, claim 0% instead of 100%, since it probably means that coverage checking has failed completely 2007-05-01 11:07:28 -07:00
Zooko O'Whielacronx
75e0284471 v0.2.0b2-0-UNSTABLE 2007-05-01 15:46:19 -07:00
Zooko O'Whielacronx
40e1c4be71 grand unified version numbers
Now the allmydata python package, the setup.py script, and the debian packages all get their tahoe version number from the same place.
2007-05-01 13:23:15 -07:00
Zooko O'Whielacronx
e697c84df1 bump timeout up because it timed out even on my super fast MacBook Pro
Hm.  This probably means that it is never going to finish...
2007-04-30 23:16:06 -07:00
Zooko O'Whielacronx
f724719744 v0.2.0b1-0-UNSTABLE 2007-04-30 23:11:23 -07:00
Zooko O'Whielacronx
137a427a47 bump some unit tests up to very high timeouts because my poor G4 867 MHz PowerBook, which is busy doing video iChat, takes a long time to run these tests 2007-04-30 21:03:43 -07:00
Brian Warner
461d64e371 test_system: increase timeouts to almost 20 minutes: we have buildslaves that take more than 5 minutes to finish these tests. 2007-04-30 20:14:58 -07:00
Zooko O'Whielacronx
e0a18d12af globally search and replace "mesh" with "grid" and adjust description of the effect of NAT on the topology 2007-04-30 13:06:09 -07:00
Zooko O'Whielacronx
b027ee1489 v0.1.4b2-0-UNSTABLE 2007-04-30 10:00:07 -07:00
Zooko O'Whielacronx
d19d1058e0 extirpate all references the "queen" and "metatracker"
This is a potentially disruptive and potentially ugly change to the code base,
because I renamed the object that serves in both roles from "Queen" to
"IntroducerAndVdrive", which is a bit of an ugly name.

However, I think that clarity is important enough in this release to make this
change.  All unit tests pass.  I'm now darcs recording this patch in order to
pull it to other machines for more testing.
2007-04-30 09:57:52 -07:00
Zooko O'Whielacronx
7090ed66c5 v0.1.3-0-UNSTABLE 2007-04-30 06:02:43 -07:00
Zooko O'Whielacronx
a61178f425 raise the limit on the number of hashes when retrieving them, too 2007-04-30 06:00:10 -07:00
Zooko O'Whielacronx
8e4aa3e1fb raise the limit on block hashes and share hashes from 30 to 2**20 2007-04-29 23:51:15 -07:00
Zooko O'Whielacronx
2fd7b81fba v0.1.2-0-UNSTABLE 2007-04-29 23:00:08 -07:00
Zooko O'Whielacronx
85b1a4374b allmydata-tahoe v0.1.1-1-UNSTABLE 2007-04-27 16:07:22 -07:00
Brian Warner
82b3554c2f testutil.py: remove unused import, appease pyflakes 2007-04-27 08:11:34 -07:00
Zooko O'Whielacronx
a70a599249 make it print out version numbers when it constructs a Node instance 2007-04-27 13:47:38 -07:00
Zooko O'Whielacronx
7b0db1c91f make runner find twistd and invoke it with python
This makes allmydata-tahoe work on Windows, and probably doesn't make it fail on other platforms.
2007-04-26 23:14:13 -07:00
Zooko O'Whielacronx
5121a8d4a2 make testutil more flexible about what platform it runs on 2007-04-26 22:16:48 -07:00
Brian Warner
db2caf58d4 interfaces: allow URIs to be up to 300 chars long, we're just crossing over the edge now 2007-04-26 18:08:29 -07:00
Brian Warner
5ceeaaea6a add RIClient.get_versions, in the hopes of enabling backwards-compatibility code in the future 2007-04-26 12:01:25 -07:00
Zooko O'Whielacronx
0df42ed71f tahoe v0.1.1-0-UNSTABLE 2007-04-26 11:25:15 -07:00
Brian Warner
4b2298937b use real encryption, generate/store/verify verifierid and fileid 2007-04-25 17:53:10 -07:00
Brian Warner
f4beff38c4 sort coverage tables by lines uncovered, rather than percentages 2007-04-24 11:49:26 -07:00
Brian Warner
9251b6e85e test_system: improve webish.py coverage 2007-04-24 01:41:54 -07:00
Brian Warner
73bb4f80af vdrive.py: let put_file return the URI that was used for the file itself 2007-04-24 01:41:40 -07:00
Brian Warner
813fdd70da test_runner.py: improve test coverage a little bit 2007-04-23 21:28:19 -07:00
Brian Warner
e7a41aa9f7 testutil.py: hush a pyflakes warning 2007-04-23 21:21:50 -07:00
Brian Warner
d1a848e0a7 runner.py: add --quiet, use it from test cases 2007-04-23 21:17:13 -07:00
Brian Warner
8307aaccb6 testutil: make SignalMixin actually be a mixin (and not inherit from TestCase), use it from all tests that start notes and thus exec ifconfig 2007-04-23 21:15:02 -07:00
Brian Warner
78430d3124 test_upload.py: add some TODO items 2007-04-23 17:30:40 -07:00
Brian Warner
491f96258e test_runner.py: add some coverage for allmydata.scripts.runner, to create nodes 2007-04-19 18:56:45 -07:00
Brian Warner
be8eeadb7a rename queen control files to 'introducer' 2007-04-19 18:24:15 -07:00
Brian Warner
a3c9ead74f replace 'queen' with 'introducer' in a lot of places, but not all 2007-04-19 17:30:21 -07:00
Brian Warner
76e0ec3214 figleaf: add number-of-uncovered-lines to the HTML output 2007-04-19 11:09:33 -07:00
Zooko O'Whielacronx
20451e43b5 assign version number tahoe 0.1.0-0-UNSTABLE 2007-04-19 13:47:59 -07:00
Zooko O'Whielacronx
be8744caf8 copy version.py from pyutil 2007-04-19 13:47:36 -07:00
Brian Warner
49e992b8b6 make test_encode less CPU-intense by using 4-out-of-10 encoding instead of 25-out-of-100 2007-04-19 10:56:15 -07:00
Brian Warner
073333c791 iputil/testutil: fix pyflakes errors/warnings 2007-04-18 18:33:37 -07:00
Brian Warner
2d0e240466 encode: handle uploads of the same file multiple times. Unfortunately we have to do almost as much work the second time around, to compute the full URI 2007-04-18 18:29:10 -07:00
Brian Warner
d1c00bd527 test_system: turn off test_connections, since it is slow and subsumed by the other system tests 2007-04-18 16:06:57 -07:00
Brian Warner
bc8ff33cd1 interfaces.py: remove spurious 'pass' statements (which, incidentally, were counted as uncovered code) 2007-04-18 15:46:37 -07:00
Brian Warner
91d72bb504 test_storage: test StorageServer code too: allocation, lookup, multiple-writer behavior 2007-04-18 15:42:34 -07:00
Brian Warner
0a53fb00b2 storageserver.py: remove unused import 2007-04-18 10:43:33 -07:00
Zooko O'Whielacronx
a154641462 iputil.list_async_addresses now "works" on cygwin 2007-04-18 17:30:08 -07:00
Zooko O'Whielacronx
47bdfdf419 copy testutil from pyutil 2007-04-18 17:28:36 -07:00
Zooko O'Whielacronx
c5b9d2929d copy repeatable_random from pyutil 2007-04-18 17:27:33 -07:00
Zooko O'Whielacronx
55949faecc finish renaminf "fec" to "zfec" in tahoe 2007-04-18 17:25:16 -07:00
Brian Warner
85b36e348b encode.py: remove unused pad() code 2007-04-17 21:22:32 -07:00
Brian Warner
80cf789817 download: remove unused import 2007-04-17 21:11:20 -07:00
Brian Warner
b76aa1ce17 download: oops, NotEnoughHashesError comes from hashtree, not hashutil 2007-04-17 20:37:51 -07:00
Brian Warner
b84d6ed07f encode: fix multi-segment uploads: lambdas inside for loops require special attention to make sure you are capturing the *value* of the loop variable and not just the slot it lives in 2007-04-17 20:29:08 -07:00
Brian Warner
a4d7607a9e test_system: minor comment 2007-04-17 20:28:19 -07:00
Brian Warner
e7ec4ff4e5 factor out the tagged hash function used for subshares/blocks 2007-04-17 20:27:56 -07:00
Brian Warner
76e28b3484 comment out some verbose log messages, add commented-out new ones 2007-04-17 20:25:52 -07:00
Brian Warner
008e418523 storageserver: assert that blocks are written in-order, clean up tests a bit 2007-04-17 20:21:05 -07:00
Zooko O'Whielacronx
c2dfcb1f01 don't clobber existing storefile every put block 2007-04-17 20:14:44 -07:00
Zooko O'Whielacronx
723f4078a3 pyutil: fileutil.open_or_create() 2007-04-17 20:14:26 -07:00
Brian Warner
e040b85f5d test_storage: add (failing) test of the BucketWriter/BucketReader implementation 2007-04-17 20:03:44 -07:00
Brian Warner
dbbbc96cbf storageserver: the last segment is not always going to be the same size as the rest, so don't assert such a thing 2007-04-17 13:41:45 -07:00
Brian Warner
2e15c9aed2 system_test: exercise multiple segments 2007-04-17 13:40:47 -07:00
Brian Warner
c3268ca394 download.py: don't truncate tail segments that are the same size as all the others 2007-04-17 13:39:35 -07:00
Brian Warner
a0dc26ee11 test_encode.Roundtrip: cover more combinations of data size relative to segment size and number of block hash tree leaves 2007-04-17 12:57:55 -07:00
Brian Warner
96812507a0 test_encode.Encode: cover more combinations of data size relative to segment size and number of block hash tree leaves 2007-04-17 12:29:56 -07:00
Brian Warner
a05b713076 test_encode: test filesizes which are an exact multiple of the segment size. This test fails right now. 2007-04-16 19:55:03 -07:00
Brian Warner
ff8cb4d32e encode: make MAX_SEGMENT_SIZE controllable, to support tests which force the use of multiple segments. Also, remove not-very-useful upload-side debug messages 2007-04-16 19:29:57 -07:00
Brian Warner
b9624502c9 download: more test coverage 2007-04-16 17:21:37 -07:00
Brian Warner
6bdabd2cea download: remove some leftover (and not very useful) debug logging 2007-04-16 17:17:57 -07:00
Brian Warner
2f5fb51848 download: validate handling of missing sharehashes too 2007-04-16 17:15:44 -07:00
Brian Warner
a5a30d6bee iputil.py: remove unused import 2007-04-16 17:08:00 -07:00
Brian Warner
42179e5ae2 download: verify that bad blocks or hashes are caught by the download process 2007-04-16 16:30:21 -07:00
Zooko O'Whielacronx
5b3e923093 storageserver: ignore files in verifierdir whose filenames aren't of the right form for shares 2007-04-18 07:41:56 -07:00
Zooko O'Whielacronx
f4de079ba0 pyutil: iputil: fix netbsd, irix, sunos 2007-04-18 07:40:26 -07:00
Brian Warner
281a0d376c test_system: bump up timeout again, the new extra download forms take more time on poor overloaded slave1 2007-04-16 15:53:25 -07:00
Brian Warner
88a7fdcaab test_iputil: improve error message 2007-04-16 15:05:25 -07:00
Brian Warner
2fef5dac1f download: log more information when hashtree checks fail 2007-04-16 13:08:19 -07:00
Brian Warner
7dabb68a51 download: improve test coverage on our IDownloadTarget classes, make FileHandle return the filehandle when its done so that it is easier to close 2007-04-16 13:07:36 -07:00
Brian Warner
652e8a735f interfaces: use explicit TupleOf and ChoiceOf constraints, since the upcoming version of Foolscap changes the meaning of bare tuples (from ChoiceOf to TupleOf) 2007-04-13 19:04:38 -07:00
Brian Warner
03fbee6ade encode.py: remove an unused import 2007-04-12 20:09:32 -07:00
Brian Warner
0ecb26ec44 test_hashtree.py: get full coverage for hashtree.py 2007-04-12 19:58:13 -07:00
Brian Warner
30133a7cdf hash trees: further cleanup, to make sure we're validating the right thing
hashtree.py: improve the methods available for finding out which hash nodes
 are needed. Change set_hashes() to require that every hash provided can
 be validated up to the root.
download.py: validate from the top down, including the URI-derived roothash
 in the share hash tree, and stashing the thus-validated share hash for use
 in the block hash tree.
2007-04-12 19:41:48 -07:00
Brian Warner
d351cd7674 download: always validate the blockhash, and don't let the bucket trick us into not validating hashes 2007-04-12 15:18:46 -07:00
Brian Warner
2e314ad47f hashtree.py: reindent from 2-spaces to 4-spaces. No functional changes. 2007-04-12 14:24:11 -07:00
Brian Warner
d8215e0c6f rename chunk.py to hashtree.py 2007-04-12 13:13:25 -07:00
Brian Warner
8f58b30db9 verify hash chains on incoming blocks
Implement enough of chunk.IncompleteHashTree to be usable.
Rearrange download: all block/hash requests now go through
a ValidatedBucket instance, which is responsible for retrieving
and verifying hashes before providing validated data. Download
was changed to use ValidatedBuckets everywhere instead of
unwrapped RIBucketReader references.
2007-04-12 13:07:40 -07:00
Zooko O'Whielacronx
234951041f oops -- the previous commit of iputil wasn't the right version
Too bad synchronizing pyutil and allmydata.util includes a manual step.
2007-04-16 15:12:01 -07:00
Zooko O'Whielacronx
ac4d23d336 port iputil to Windows (and Irix, and NetBSD, and Solaris 2, ...) 2007-04-16 14:59:13 -07:00
Brian Warner
64ea644a48 test_util: add more coverage for assertutil.py 2007-04-08 13:02:13 -07:00
Brian Warner
6f4280cc38 test_util: add full coverage for mathutil.py 2007-04-08 12:43:01 -07:00
Brian Warner
d8b71b85f8 download: retrieve share hashes when downloading. We don't really do much validation with them yet, though. 2007-04-06 22:51:19 -07:00
Brian Warner
9a64a9f76e codec: remove more unused debugging code 2007-04-06 20:56:39 -07:00
Brian Warner
5b0ece6ece node: enable Tub.logLocalFailures now that we require foolscap-0.1.2 2007-04-06 20:55:59 -07:00
Brian Warner
245d318409 codec: remove some unused debugging code 2007-04-06 20:55:37 -07:00
Brian Warner
fe786350dc test_system: bump up timeouts for the sake of slow slave1, give each test a separate base directory 2007-04-06 20:34:32 -07:00
Brian Warner
5168c7c8d6 encode: add more logging to investigate occasional test failures 2007-04-06 18:04:38 -07:00
Brian Warner
dd673370ec figleaf output: include a stats.out for parsing by a munin plugin 2007-04-06 17:41:01 -07:00
Brian Warner
6a6631a92e test_util.py: sigh, one last minor python-2.5 issue 2007-04-06 17:21:25 -07:00
Brian Warner
736fbe6ea5 test_util.py: fix another minor python-2.5 issue 2007-04-06 17:12:26 -07:00
Brian Warner
e74e370d00 test_util.py: fix problems 2007-04-06 16:36:22 -07:00
Brian Warner
d432099880 more allmydata.util test coverage, rename the test case to be more correct 2007-04-06 16:33:22 -07:00
Brian Warner
1a4cf6f7b7 add test coverage for stuff in allmydata/utils 2007-04-06 16:29:25 -07:00
Brian Warner
97a40bf20f encode/upload: add more logging, to understand the test failure on a slow buildslave 2007-04-06 15:45:45 -07:00
Brian Warner
654854635a chunk: add IncompleteHashTree for download purposes, plus tests 2007-04-06 09:09:57 -07:00
Brian Warner
8d2def5b04 encode: clean up some weirdness that was there to make unit tests easier to write 2007-04-05 22:36:18 -07:00
Brian Warner
919ca3e902 rename encode_new.py to encode.py, now that there isn't an old one anymore 2007-04-05 21:17:42 -07:00
Brian Warner
2122fbaca9 tests: clean up tearDown to use flushEventualQueue instead of hacking fixed-time delays 2007-04-04 16:09:13 -07:00
Brian Warner
92ba8dc8d2 allmydata.interfaces: remove some of the placeholders now that we require foolscap-0.1.2 2007-04-04 15:59:36 -07:00
Zooko O'Whielacronx
223bf98992 use the "binary" flag on open() for files that shouldn't have line-endings automatically converted 2007-04-04 16:12:30 -07:00
Zooko O'Whielacronx
cd4472003b tidiness: return res from logging of progress 2007-04-02 10:23:24 -07:00
Brian Warner
8bc248a515 remove util.ring, no longer used. also remove test_ring.py 2007-03-31 20:38:43 -07:00
Zooko O'Whielacronx
d5be3e6086 fix BucketWriter to not create a finalhome until it is complete, and to clean up the empty $VERIFIERID dir under incoming/ when it moves the last share out of it 2007-03-30 18:01:56 -07:00
Zooko O'Whielacronx
82f7303179 log a running total of how much of your file has been uploaded 2007-03-30 18:01:37 -07:00
Zooko O'Whielacronx
077eb7507c assert that only dicts get passed to _got_response() 2007-03-30 18:00:40 -07:00
Zooko O'Whielacronx
1a20e5d01b fix storage server to handle the case that it has no directory at all when someone asks for buckets 2007-03-30 17:12:07 -07:00
Zooko O'Whielacronx
826ce43be7 fix bug in storage-server: yes, "0" is a number, Mr. storage server 2007-03-30 17:07:04 -07:00
Brian Warner
466bb46abe handle foolscap's annoying tendency to turn everything into a sets.Set, which are not interoperable with 2.4's builtin 'set' type 2007-03-30 16:55:04 -07:00
Brian Warner
7cd9ef3bbf finish making the new encoder/decoder/upload/download work 2007-03-30 16:50:50 -07:00
Zooko O'Whielacronx
3d694a90f3 add unit tests and fix bugs in upload 2007-03-30 14:54:33 -07:00
Brian Warner
234b2f354e add new test for doing an encode/decode round trip, and make it almost work 2007-03-30 13:20:01 -07:00
Brian Warner
2593ce42c3 uri.py: share counts are not base32-encoded 2007-03-30 12:36:16 -07:00
Brian Warner
f5a49aa7ee upload: change _compute_uri a bit, get infile in a different way 2007-03-30 12:30:14 -07:00
Brian Warner
9a2e0cf28e switch upload to use encode_new, fix a few things (but not nearly all of them) 2007-03-30 11:53:03 -07:00
Brian Warner
7d7ce7a059 test_encode: make sure encode_new can produce the data it is supposed to 2007-03-30 11:32:57 -07:00
Brian Warner
f55b9814ec chunk.py: remove unused non-tagged hash code 2007-03-30 11:32:47 -07:00
Brian Warner
32512b73d6 encode_new: use tagged (sha256) hashes everywhere 2007-03-30 11:32:13 -07:00
Zooko O'Whielacronx
62456b2a01 import fileutil, some of which came from amdlib.util and some of which came from the pyutil library 2007-03-30 12:12:23 -07:00
Zooko O'Whielacronx
f4a718c5b6 finish storage server and write new download 2007-03-30 10:52:19 -07:00
Zooko O'Whielacronx
17299fc96e new upload and storage server 2007-03-29 20:19:52 -07:00
Brian Warner
83fcda82f6 chunk.py: fix a pyflakes warning 2007-03-29 19:49:40 -07:00
Brian Warner
3376c2b0a1 node creation: be willing to utilize a pre-existing (but empty) directory 2007-03-29 14:32:28 -07:00
Brian Warner
60bddba827 webish: add 'my nodeid' to the page 2007-03-29 14:31:55 -07:00
Zooko O'Whielacronx
c70b4773d3 change HashTree to use new hashutil convenience methods, thus fixing a security flaw 2007-03-29 18:12:35 -07:00
Zooko O'Whielacronx
99a046ab51 hashutil: convenience methods for tagged and encoded hashes
In various cases, including Merkle Trees, it is useful to tag and encode the inputs to your secure hashes to prevent security flaws due to ambiguous meanings of hash values.
2007-03-29 18:11:30 -07:00
Zooko O'Whielacronx
4b4f5bbcba change #!/usr/bin/python to #!/usr/bin/env python
Note that using "whatever version of python the name 'python' maps to in the current shell environment" is more error-prone that specifying which python you mean, such as by executing "/usr/bin/python setup.py" instead of executing "./setup.py".  When you build tahoe (by running "make") it will make a copy of bin/allmydata-tahoe in instdir/bin/allmydata-tahoe with the shebang line rewritten to execute the specific version of python that was used when building instead of to execute "/usr/bin/env python".

However, it seems better that the default for lazy people be "whatever 'python' means currently" instead of "whatever 'python' meant to the manufacturer of your operating system".
2007-03-29 14:01:28 -07:00
Brian Warner
93c4a5ebb0 test_iputil: remove the test that only works on linux, since we're using the cross-unix 'get_local_addresses_async' anyways. This should allow the tests to pass on OS-X 2007-03-29 11:21:17 -07:00
Brian Warner
3c757a524f introducer: add some comments about separating nodeid from tubid 2007-03-29 11:16:29 -07:00
Brian Warner
7802aa5303 hush pyflakes warnings in codec and test_codec 2007-03-28 15:31:51 -07:00
Brian Warner
6f718d72a9 encode_new.py: rearrange methods into the order in which they should be called 2007-03-28 13:30:17 -07:00
Brian Warner
74d0839980 update memory-footprint and share-size docs in ICodecEncoder 2007-03-28 12:52:44 -07:00
Brian Warner
dc25f327a3 document a proposed IDecoder interface, still pretty bare-bones 2007-03-28 11:31:31 -07:00
Brian Warner
c4f09833f5 document IEncoder, add Encoder.set_landlords() 2007-03-28 11:24:53 -07:00
Brian Warner
a159ab3be8 encode_new.py: add comments, make variable names more illuminating 2007-03-28 11:06:19 -07:00
Zooko O'Whielacronx
0a9a6f0216 remove left-over early encode mechanism 2007-03-28 00:06:03 -07:00
Zooko O'Whielacronx
81c8cb6d6a switch to pyfec 2007-03-28 00:05:16 -07:00
Zooko O'Whielacronx
e4463056f3 fix test_codec and test_upload to handle current API 2007-03-27 22:57:15 -07:00
Brian Warner
0ee2f0f3bf update ReplicatingEncoder and decoder to match the current ICodecEncoder/decoder interface 2007-03-27 22:17:31 -07:00
Brian Warner
2b2e0fc0d4 ICodecEncoder: resolve some questions about the API, still more to examine 2007-03-27 20:14:45 -07:00
Brian Warner
59201843b6 update the docs on ICodecEncoder and ICodecDecoder 2007-03-27 19:05:09 -07:00
Brian Warner
611178490a fix some python2.5 incompatibilities, and remove an old webish display that suggested we might know about peers but not be connected to them 2007-03-27 17:44:49 -07:00
Brian Warner
51ad4418f7 test_introducer: flushEventualQueue at the end of the test run 2007-03-27 17:16:13 -07:00
Brian Warner
25ff9e1f97 complete the Introducer changes, separate out vdrive access, make everything work again 2007-03-27 16:12:11 -07:00
Zooko O'Whielacronx
b58aa3c371 merge incomplete stuff with other patches 2007-03-23 16:20:26 -07:00
Zooko O'Whielacronx
e48a8b8e66 incomplete work to be finished elsewhere 2007-03-23 16:15:57 -07:00
Brian Warner
1264d77fd5 hush a pyflakes warning 2007-03-22 22:22:00 -07:00
Zooko O'Whielacronx
c77eace162 separate queen's service of introduction-to-the-network from queen's service of providing-access-to-the-vdrive 2007-03-22 14:39:30 -07:00
Brian Warner
ca33c8f848 more work on a memory-footprint test program 2007-03-12 16:28:37 -07:00
Brian Warner
1e25c9e897 start work on a memory-measuring test tool 2007-03-08 18:12:24 -07:00
Brian Warner
e08308e019 test_system.py: remove the lowered (20s) timeouts, since some buildslaves require more like 30 or 40 seconds to complete the test 2007-03-08 17:30:09 -07:00
Brian Warner
b9c143b0e8 test_observer: improve test coverage 2007-03-08 16:13:14 -07:00
Brian Warner
91e7f092e8 node.py: switch to using get_local_addresses_async, which is slightly more portable (most forms of unix) 2007-03-08 15:12:52 -07:00
Brian Warner
78d19c271c rearrange service startup a bit, now Node.startService() returns a Deferred that fires when the tub is actually ready, and there is also a Node.when_tub_ready() hook. This allows get_local_addresses() to be slow and not break everything. Changed all necessary test cases to accomodate this slow startup. 2007-03-08 15:10:36 -07:00
Brian Warner
2f02659b5a test_observer.py: remove the code that allows observers to run out-of-order, our use of foolscap.eventual removes this error case 2007-03-08 15:09:29 -07:00
Brian Warner
929d725577 add OneShotObserverList from the amdlib tree 2007-03-08 15:07:38 -07:00
Brian Warner
228e17560a control.py: fix get_memory_usage, add a sample client tool 2007-03-07 20:31:49 -07:00
Brian Warner
751587f376 iputil: switch to a scapy-inspired SIOCGIFADDR approach, very linux-specific now 2007-03-07 20:03:47 -07:00
Brian Warner
eccb3258ae iputil: use the subprocess module instead of os.popen 2007-03-07 19:29:00 -07:00
Brian Warner
d4020fa533 control: add RIControlClient, fix some pyflakes warnings 2007-03-07 19:20:27 -07:00
Brian Warner
a71d83bd8e add a local foolscap control interface, to upload/download files and check memory usage 2007-03-07 19:16:06 -07:00
Brian Warner
b5bf2baa9b iputil: use explicit /sbin/ifconfig, to avoid depending upon PATH 2007-03-07 18:47:40 -07:00
Brian Warner
2c261ce996 change node startup to put all local addresses in the PBURL, including 127.0.0.1. This should facilitate testing on both connected and disconnected systems. 2007-03-07 18:43:17 -07:00
Brian Warner
18325251bf iputil: add get_local_addresses(), an attempt to enumerate all IPv4 addresses on this host. This is pretty unix-specific for right now (it calls ifconfig) 2007-03-07 18:22:30 -07:00
Brian Warner
f2956c257b util.iputil: try to survive not having a global network connection at all 2007-03-07 18:21:42 -07:00
Brian Warner
61760047cf add some questions to allmydata.interfaces 2007-03-05 20:57:38 -07:00
Brian Warner
9d123605cb hush pyflakes warnings 2007-02-01 17:13:01 -07:00
Zooko O'Whielacronx
c427b880d2 update the use of the encoder API in download.py 2007-02-01 16:30:13 -07:00
Zooko O'Whielacronx
dd4ad3d542 use pyfec instead of py_ecc for erasure coding and update API to codec 2007-02-01 16:07:00 -07:00
Zooko O'Whielacronx
1373789463 add mathutil.next_power_of_k() and mathutil.ave() 2007-02-01 15:55:26 -07:00
Brian Warner
e4870a432c fix some of the filetree/workqueue interface definitions 2007-01-26 17:31:24 -07:00
Brian Warner
ef73ebaf0a download: update all users to match Zooko's change to ICodecDecoder.decode (as it now returns a list instead of a single string) 2007-01-24 17:23:22 -07:00
Zooko O'Whielacronx
833ef86a88 update interfaces and docs for codec
It now takes a sequence of buffers instead of a single string for both encode and decode, and it also takes a separate sequence of shareids for decode instead of a sequence of tuples, and it returns a sequence of buffers instead of a single string.
2007-01-24 15:34:02 -07:00
Brian Warner
c27564a3e2 hush pyflakes 2007-01-24 15:29:22 -07:00
Brian Warner
bbf188d2c6 filetree: make delete() work 2007-01-24 15:10:53 -07:00
Brian Warner
22731125f3 filetree: change the way addpath works, now we add workqueue steps for all involved subtrees at about the same time, rather than letting one step add the next when it runs. Finally add a (passing) test for uploading files to CHK-based directories 2007-01-22 01:06:09 -07:00
Brian Warner
9dc1c0cfc0 filetree: add vdrive upload/download test, change workqueue relative-filename semantics 2007-01-21 16:03:15 -07:00
Brian Warner
5453e0f022 queen.sendOnly: ignore DeadReferenceError too 2007-01-21 16:01:09 -07:00
Brian Warner
81d093b649 move IWorkQueue into allmydata.interfaces, give VirtualDrive an uploader 2007-01-21 15:15:31 -07:00
Brian Warner
430b3a03fc move upload/download interfaces to allmydata.interfaces, let SubTreeMaker assert IDownloader-ness of its 'downloader' argument 2007-01-21 15:01:34 -07:00
Brian Warner
a8ecaf45b6 test_filetree: more vdrive._get_file_uri() coverage 2007-01-21 13:39:40 -07:00
Brian Warner
124d531b41 filetree: mark leaf nodes by adding is_leaf_subtree(), stop traversing when we hit them, to make vdrive._get_file_uri() work 2007-01-21 13:31:16 -07:00
Brian Warner
9e7dbf20bc filetree: test NoSuchDirectoryError in vdrive.list() 2007-01-21 04:50:37 -07:00
Brian Warner
7b8c524d7c more filetree, workqueue-boxes now hold serialized Nodes, move NodeMaker out to a separate module 2007-01-21 04:18:54 -07:00
Brian Warner
324033c9e2 filetree: start testing IVirtualDrive, beginning with list() 2007-01-20 21:14:41 -07:00
Brian Warner
8c7d33f4a2 filetree: put SubTreeMaker and NodeMaker in separate classes 2007-01-20 17:04:56 -07:00
Brian Warner
ce4610c3e6 more filetree work, more tests now pass 2007-01-20 15:50:21 -07:00
Brian Warner
b61a4ff371 more filetree, it's actually starting to make sense now 2007-01-20 14:41:51 -07:00
Brian Warner
48216ce272 filetable: oops, os.listdir() does not guarantee sorted results 2007-01-20 13:02:26 -07:00
Brian Warner
3d76250894 filetree: more tests, still very early 2007-01-20 05:13:15 -07:00
Brian Warner
c808d5a5ef filetree: refactor INode serialization, start on tests 2007-01-20 04:52:53 -07:00
Brian Warner
e843abe542 more filetree hacking, still too early to test 2007-01-20 04:25:20 -07:00
Brian Warner
4a0f8bc110 checkpont more filetree stuff 2007-01-19 23:22:39 -07:00
Brian Warner
8921c1b666 snapshot filetree work: fix pyflakes complaints 2007-01-19 02:35:36 -07:00
Brian Warner
ff6b09d973 snapshot filetree work: it's getting close 2007-01-19 02:23:03 -07:00
Brian Warner
ca3fda3e22 download.py: fix IDownloader to take a URI 2007-01-19 02:17:48 -07:00
Brian Warner
eeffcda7b1 filetree.interfaces: remove not-really-code to improve coverage stats 2007-01-17 17:00:47 -07:00
Brian Warner
56df1a48f4 improve test coverage a bit 2007-01-17 15:34:29 -07:00
Brian Warner
fcf5721935 webish: present real URI in directory listing, not an unnecessarily b2a'ed form 2007-01-17 14:46:02 -07:00
Brian Warner
c94098b93a split filetree_new.py up into smaller pieces, in a new subpackage 2007-01-17 13:54:38 -07:00
Brian Warner
18ec38acf0 hush pyflakes warnings 2007-01-16 21:34:34 -07:00
Brian Warner
4101bcf218 update URI format, include codec name 2007-01-16 21:29:59 -07:00
Brian Warner
8e03d19ccc webish: add PBURL to the all-peers table on the welcome page 2007-01-16 21:01:18 -07:00
Brian Warner
c868f77c71 encode_new.py: recent Foolscap accepts 'None' as a constraint 2007-01-16 19:56:30 -07:00
Brian Warner
3fb37ae246 webish: more verifierid-to-uri transition 2007-01-16 19:55:53 -07:00
Brian Warner
43f32e8c9b vdrive/webish: finish verifierid/uri transition 2007-01-16 19:43:13 -07:00
Brian Warner
453f0b47e4 upload: fix typo in debug messages 2007-01-16 19:42:28 -07:00
Brian Warner
ccdc2622d4 upload: rearrange peer-selection code to be more readable, and fix a silly bug 2007-01-16 19:35:12 -07:00
Brian Warner
460ed84562 oops, fix that NotEnoughPeersError instrumentation 2007-01-16 19:07:23 -07:00
Brian Warner
fed539a16d add better error reporting to NotEnoughPeersError 2007-01-16 18:47:52 -07:00
Brian Warner
8d4d2c82c4 change upload to push 2 shares instead of 3
Now that peers can talk to themselves, the 5-node system test won't fail
just because one of the shares was hosted on the downloader (and thus
inaccessible until recently). The 3-share push was there to avoid this
problem.
2007-01-16 15:15:26 -07:00
Brian Warner
56cf2dbd7e now that foolscap-0.0.7 accepts connect-to-self, allow peers to know about themselves. We now require foolscap-0.0.7 2007-01-16 15:12:49 -07:00
Brian Warner
2150ac0e53 hush pyflakes warnings 2007-01-15 21:31:10 -07:00
Brian Warner
3209fd5e09 rearrange encode/upload, add URIs, switch to ReplicatingEncoder
Added metadata to the bucket store, which is used to hold the share number
(but the bucket doesn't know that, it just gets a string).

Modified the codec interfaces a bit.

Try to pass around URIs to/from download/upload instead of verifierids.
URI format is still in flux.

Change the current (primitive) file encoder to use a ReplicatingEncoder
because it provides ICodecEncoder. We will be moving to the (less primitive)
file encoder (currently in allmydata.encode_new) eventually, but for now
this change lets us test out PyRS or zooko's upcoming C-based RS codec in
something larger than a single unit test. This primitive file encoder only
uses a single segment, and has no merkle trees.

Also added allmydata.util.deferredutil for a DeferredList wrapper that
errbacks (but only when all component Deferreds have fired) if there were
any errors, which unfortunately is not a behavior available from the standard
DeferredList.
2007-01-15 21:22:22 -07:00
Brian Warner
78a9e815c5 add simple metadata (a single string) to the storage protocol 2007-01-15 14:01:22 -07:00
Brian Warner
e29032bc36 change IEncoder to ICodecEncoder, to match the previous change 2007-01-11 20:57:14 -07:00
Brian Warner
417c17755b use the word 'codec' for erasure coding, for now. 'encode' is used for file-level segmentation/hashing 2007-01-11 20:51:27 -07:00
Brian Warner
a65608fec1 sendOnly: oops, I keep forgetting that you can't really use f.trap in a lambda, because it returns the failure it trapped 2007-01-09 21:13:05 -07:00
Brian Warner
ceda350892 test_system.py: exercise queen.Roster._lost_node too 2007-01-09 19:40:36 -07:00
Brian Warner
aad0a9dfac workqueue: more tests 2007-01-09 01:58:50 -07:00
Brian Warner
d61c0a6ef6 workqueue: more tests, coverage now at 63.4%, yay 2007-01-09 01:36:12 -07:00
Brian Warner
b641f6cbc7 workqueue: more improvements, more tests 2007-01-08 22:29:42 -07:00
Brian Warner
b9edb02820 workqueue: start adding tests 2007-01-08 21:29:33 -07:00
Brian Warner
2567a26519 rename the new filetable code to 'filetree', since robk astutely pointed out that 'table' is misleading and implies a flat list of files in a single directory 2007-01-08 16:29:14 -07:00
Brian Warner
5a34753cdb fix pyflakes warning in debugshell, by providing a dummy value for 'app' that will be overwritten when the manhole connection is established 2007-01-05 21:15:59 -07:00
Brian Warner
9382d61208 use absolute import of 'allmydata.Crypto' rather than a relative import of just 'Crypto', to make it clear that we're using our own form rather than relying upon the system version 2007-01-05 21:12:26 -07:00
Brian Warner
94b1840295 workqueue: fix pyflakes warnings, code is still quite incomplete 2007-01-05 21:11:19 -07:00
Brian Warner
2566452091 test_encode_share.py: fix some pyflakes warnings 2007-01-05 18:53:22 -07:00
Brian Warner
4ba6f47d2c add some (disabled) encoder benchmarking code 2007-01-05 18:46:03 -07:00
Brian Warner
cf04073e98 change Encoder to use py_ecc, now that it works 2007-01-05 18:12:45 -07:00
Brian Warner
42c0d2e336 disable figleaf tracing during py_ecc, since it takes *forever*, especially on the slow buildslave 2007-01-05 18:12:04 -07:00
Brian Warner
3d3a7a5b8d encode.py: add some timing comments 2007-01-05 00:48:42 -07:00
Brian Warner
409d92e746 only run a single (short) py_ecc test on slave3, since it is so slow the tests timeout 2007-01-05 00:42:52 -07:00
Brian Warner
e1c6ee9dcf encoding: fix the last py_ecc problem, tests pass now 2007-01-05 00:06:42 -07:00
Brian Warner
436409fa4d change encode_new to use IEncoder 2007-01-04 23:51:35 -07:00
Brian Warner
c91d14dca8 fix our use of py_ecc (set log2FieldSize=8 explicitly) 2007-01-04 23:50:21 -07:00
Brian Warner
f31fc06d89 establish IEncoder/IDecoder, create suitable interfaces for both the simple replicating encoder and the py_ecc one, add a (failing) unit test for it 2007-01-04 21:52:51 -07:00
Brian Warner
fa11c3f722 more pyflakes cleanups 2007-01-04 18:06:20 -07:00
Brian Warner
6587645f76 fix pyflakes warnings in figleaf 2007-01-04 18:04:43 -07:00
Zooko O'Whielacronx
5ad6c43622 add padding to segments and convert type from string to list of integers before passing to py_ecc
But then I get an error that I don't understand about ECC math...
2007-01-04 17:58:14 -07:00
Brian Warner
7b2268ab39 improve figleaf: fix some line-numbering bugs, add an alphabetically-sorted index HTML page 2007-01-04 01:26:43 -07:00
Brian Warner
8a78065f16 modify figleaf2html to show module names instead of .py filenames, also add a --root argument to restrict coverage to a specific parent directory 2007-01-03 22:23:25 -07:00
Brian Warner
e8ee3365ef move figleaf2html and figleaf_htmlizer.py into our tree, for easier customization 2007-01-03 22:06:51 -07:00
Brian Warner
f97eec893e figleaf: move a copy into allmydata.util.figleaf, update Makefile/trial stuff 2007-01-03 21:38:29 -07:00
Brian Warner
e600571f82 checkpoint work-in-progress for WorkQueue, a disk-persistent list of work to be done 2007-01-01 23:47:16 -07:00
Brian Warner
a6678b9a37 filetable_new: import py_ecc, make the tests get a little bit further 2007-01-01 23:44:12 -07:00
Zooko O'Whielacronx
b0315fc549 make encode_new use py_ecc for real live erasure coding
(This patch is not tested -- I'm working on a Mac which doesn't have gcc
installed...  (gcc is necessary for the crypto module.)  I will now attempt to
connect to a better development computer over my mom's staticy, failure-prone,
14.4 K modem...)
2006-12-29 13:50:53 -07:00
Zooko O'Whielacronx
f3f91105bd import mathutil from pyutil 2006-12-29 13:50:42 -07:00
Zooko O'Whielacronx
dcfddcac41 encode_new: fix a bug in subshare hashes and tidy-up a couple of things 2006-12-29 12:40:10 -07:00
Brian Warner
cf53abab42 more filetable_new tests 2006-12-25 00:56:18 -07:00
Brian Warner
3e085efa7c filetable_new: fix the test 2006-12-24 12:47:03 -07:00
Brian Warner
70f5c13e26 checkpointing new filetable work.. tests don't pass yet 2006-12-24 12:39:24 -07:00
Brian Warner
56ff0e2b57 filetable: shuffle lines a bit to appease figleaf's confusion 2006-12-14 04:33:57 -07:00
Brian Warner
eb3b9b16af encode: start to fix a few problems, still a lot of work left to go 2006-12-14 04:31:17 -07:00
Brian Warner
3e5e2ee9fe make initial simple encode_new test pass 2006-12-14 04:17:01 -07:00
Brian Warner
3490378551 move all packages into src/, fix allmydata.Crypto build. Now you must perform a 'setup.py build' before using anything, and you must add the build directory (build/lib.linux-i686-2.4) to your PYTHONPATH before doing anything 2006-12-14 03:39:50 -07:00