| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2017-12-05 | [bug] transfer maximum allowed number of blobs | drebs | |
| -- Closes: #9004 | |||
| 2017-12-04 | [bug] escape tag and size | Victor Shyba | |
| It was generating spaces, causing split to fail sometimes. | |||
| 2017-12-04 | [style] fixes from code review | Victor Shyba | |
| 2017-12-01 | [feature] stream blobs from a CooperativeTask | Victor Shyba | |
| Adds a CooperativeTask to stream blobs to client as the transport is able to handle it, pausing and resuming as necessary. -- Related: #8809 | |||
| 2017-12-01 | [feature] adds a stream downloader | Victor Shyba | |
| First version, still missing consumer/producer model and some tweaks, but working. -- Related: #8809 | |||
| 2017-12-01 | [feature] add implementation for get_blob_size | Victor Shyba | |
| 2017-12-01 | [test] init StreamingResource on test_blobs_server | Victor Shyba | |
| -- Related: #8809 | |||
| 2017-12-01 | [refactor] get namespace as an optional query arg | Victor Shyba | |
| 2017-12-01 | [bug] missing self parameter on class method | Victor Shyba | |
| 2017-12-01 | [feature] add a streaming resource | Victor Shyba | |
| -- Related: #8809 | |||
| 2017-12-01 | [style] fix typo | Victor Shyba | |
| 2017-11-30 | [style] remove public get/set blob priority methods | drebs | |
| 2017-11-30 | [bug] avoid overwriting priorities on sync status batch update | drebs | |
| 2017-11-30 | [style] some sugar for setting the priority | Kali Kaneko | |
| 2017-11-30 | [feature] add priorities for blob transfers | drebs | |
| Closes: #8691 | |||
| 2017-11-29 | [feature] add bandwidth throttling to the client | drebs | |
| Intercept the creation of the protocol factory in the HTTP connection pool to use twisted.protocols.policies.ThrottlingFactory and control the incoming and outgoing bandwidth. The factory only controls one connection, so when throttling we limit the number of connections of the pool to one per host. This way, throttling happens in a per-host basis. Closes: #8931 | |||
| 2017-11-22 | [refactor] simplify decorated test parameters | Victor Shyba | |
| 2017-11-22 | [test] improve isolation by avoiding _all_dbs | Victor Shyba | |
| Accessing _all_dbs slows down the test and open possibilites for heisenbugs where some old leftover database can change code behavior. -- Resolves: #9001 | |||
| 2017-11-22 | [refactor] extract list dbs method | Victor Shyba | |
| 2017-11-22 | [refactor] remove unused parameter | Victor Shyba | |
| 2017-11-22 | [pkg] update changelog0.10.5 | drebs | |
| 2017-11-22 | [pkg] update debian changelog | drebs | |
| 2017-11-22 | [pkg] bump sphinx release info | drebs | |
| 2017-11-22 | [test] avoid collision when running fs backend tests in parallel | drebs | |
| Closes: #9000 | |||
| 2017-11-22 | [bug] remove SOLEDAD_COUCH_URL test env | Victor Shyba | |
| It was None for parallel tests and there is already a mechanism in place for setting this URL, which avoids it being set in two different places. | |||
| 2017-11-22 | [feature] make logs less noisy by monkey-patching ↵ | drebs | |
| twisted.internet.protocol.Factory | |||
| 2017-11-22 | [bug] properly log missing schema version in config doc | drebs | |
| 2017-11-22 | [bug] ensure shutdown happens only after reactor start | drebs | |
| 2017-11-22 | [bug] fix logging while checking CouchDB schema versions on server startup | drebs | |
| 2017-11-21 | [bug] use os._exit to ensure a clean log on startup failure | drebs | |
| 2017-11-21 | [docs] add entry for next release to changelog | Kali Kaneko | |
| 2017-11-20 | [test] avoid failing when there's no couch_url in environment variable | drebs | |
| 2017-11-20 | [bug] log and exit on server startup errors | drebs | |
| Some errors during server startup could leave the server in a zombie state (running, but not listening). This commit makes sure the server stops if errors occur during deferreds created on server startup. Closes #8997. | |||
| 2017-11-20 | [bug] use reactor.stop() to exit when server startup checks fail | drebs | |
| Our current use of sys.exit(20) to stop the server when startup checks fail affects logging in a bad way. This commit uses a system event trigger to exit with the desired status code when startup checks fail. Closes: #8996 | |||
| 2017-11-20 | [test] always use a tempdir for server fixture config file | drebs | |
| When starting up a Soledad Server through the use of a fixture in tests, a config file is created with the options needed for that test. As the config file was being created in /etc and not being removed, this was influencing servers started by other tests that don't use that fixture. By always using a temporary directory in the server fixture, we make sure that the config file is used only for that test and is removed after the test is finished. Closes: #8995 | |||
| 2017-11-19 | [test] use environment to pass the couch_url to server TAC test | drebs | |
| 2017-11-19 | [feature] allow setting couchdb url from environment | drebs | |
| 2017-11-19 | [test] TAC test now checks for server (with retries) | drebs | |
| After we fixed the server to wait for checks before listening on ports (201ef7a9b979f8c8efaedbe542c631944d8956f4), the TAC test started failing randomly in our CI, probably because sometimes the server takes more than 1 second to startup in the CI setup (docker in a vm). This commit adds a check with retry, and so the test will retry 10 times with an interval of 1 second before failing. | |||
| 2017-11-19 | [benchmarks] pin elasticsearch-py version to avoid TLS certs bug | drebs | |
| elasticsearch-py 6.0.0 introduced a TLS cert verification bug: https://github.com/elastic/elasticsearch-py/issues/669 That bug affects our benchmarks ci pipeline: https://0xacab.org/leap/soledad/issues/8993#note_128667 This commit only pins the version of elasticsearch-py used by benchamarks to avoid that bug. This pinning should probably be removed when the next version of elasticsearch-py is released. Closes #8993. | |||
| 2017-11-15 | [pkg] update debian changelog0.10.4 | drebs | |
| 2017-11-15 | [pkg] bump sphinx release info | drebs | |
| 2017-11-15 | [pkg] update changelog | drebs | |
| 2017-11-14 | [refactor] factor server twisted app to it's own .py file | drebs | |
| 2017-11-14 | [style] comment fix and variable renaming | drebs | |
| 2017-11-14 | [bug] use custom logger everywhere in server | drebs | |
| 2017-11-14 | [bug] remove unneeded delayed initialization of couch state | drebs | |
| 2017-11-14 | [bug] wait for checks before running servers | drebs | |
| 2017-11-14 | [refactor] rename entrypoints to match design docs | drebs | |
| 2017-11-14 | [pkg] make debian package report the correct version | drebs | |
| The Soledad python source currently makes use of versioneer to calculate versions dynamically and hardcode versions upon creation of the python package. Unfortunatlly, this procedure does not behave well with debian packaging, and we need to manually generate/replace the _version.py file so the code that ends up in the debian package reports the correct version of the software. This commit adds a script to replace the _version.py file with a static one and rules during packaging to run that script during preparation stage of debian helper. | |||
| 2017-11-14 | [pkg] add man page for soledad-server | drebs | |
 |
index : soledad.git | |
| [soledad] | git repository hosting |
| summaryrefslogtreecommitdiff |