Age | Commit message (Collapse) | Author | |
---|---|---|---|
2017-11-03 | [doc] add blobs api doc | drebs | |
2017-11-02 | [doc] add blobs sync docs | drebs | |
2017-11-02 | [doc] split blobs doc in many pages | drebs | |
2017-11-02 | [benchmarks] make blob size configurable in scalability tests | drebs | |
2017-11-02 | [benchmarks] use only one user in scalability tests | drebs | |
2017-11-02 | [benchmarks] add some orchestration to scalability tests | drebs | |
2017-11-02 | [benchmarks] add code for stressing the server | drebs | |
2017-11-02 | [benchmarks] add tokens setup to server scalability test controller | drebs | |
2017-11-02 | [benchmarks] add blob template creator for scalability test | drebs | |
2017-11-02 | [benchmarks] add test controller for scalability tests | drebs | |
2017-10-31 | [bug] set errbacks before gathering results | Victor Shyba | |
2017-10-31 | [feat] improve speed of server startup | drebs | |
To avoid corrupting data, Soledad Server checks all user databases during startup to make sure all of them use the correct schema version. This was done synchronously, so when there are many databases startup would take a long time. This commit makes that verification asynchronous, thus speeding up server startup. | |||
2017-10-31 | [style] wrong/missing docs/comments (code review) | Victor Shyba | |
2017-10-31 | [feature] add sync progress attribute | Victor Shyba | |
--Resolves: #8848 | |||
2017-10-27 | [bug] there is no retry limit for usual transfers | Victor Shyba | |
Retry limit was originally specified in #8825 as a protection mechanism, but #8822 (retry) doesn't specify a retry limit. In fact, blobs is supposed to retry until transfer is complete using timed delays between attempts, but never giving up. -- Related: #8822 -- Related: #8825 | |||
2017-10-27 | [refactor] semaphore.run instead acquire/release | Victor Shyba | |
2017-10-27 | [refactor] add a table for sync_status | Victor Shyba | |
As defined in #8970, this table and the new module will ease adding sync features such as priority queues and streaming. --Resolves: #8970 | |||
2017-10-27 | [ci] pep8 got renamed to pycodestyle | Victor Shyba | |
see: https://github.com/PyCQA/pycodestyle/issues/466 | |||
2017-10-27 | [style] E722 do not use bare except | Victor Shyba | |
2017-10-27 | [bug] TypeError: Incorrect padding | Victor Shyba | |
preamble.py wasn't using urlsafe version of base64, while all other parts of blobs were using it. --Resolves: #8980 | |||
2017-10-27 | [refactor] split blobs into modules | Victor Shyba | |
So we can have manager, sync, sql and errors in its own places. --Related: #8970 | |||
2017-10-26 | remove platform packages for unsupported distributions | Micah Anderson | |
2017-10-26 | [pkg] Lintian: build-depends-on-obsolete-package | Varac | |
2017-10-26 | [pkg] Linitan: ignore newer-standards-version | Varac | |
2017-10-26 | [pkg] Fix lintian error | Varac | |
Fix this lintian error: 00:14:29 W: soledad source: maintainer-script-lacks-debhelper-token debian/soledad-server.postinst 00:14:29 N: 00:14:29 N: This package is built using debhelper commands that may modify 00:14:29 N: maintainer scripts, but the maintainer scripts do not contain the 00:14:29 N: "#DEBHELPER#" token debhelper uses to modify them. 00:14:29 N: 00:14:29 N: Adding the token to the scripts is recommended. 00:14:29 N: 00:14:29 N: Severity: normal, Certainty: possible 00:14:29 N: 00:14:29 N: Check: debhelper, Type: source 00:14:29 N: | |||
2017-10-26 | Standards-Version 4.1.1, Ubuntu artful package | Varac | |
2017-10-23 | [bug] revert unintentional changes from last commit | drebs | |
2017-10-23 | [doc] add script for copying doc to leap_se repo | drebs | |
2017-10-23 | [doc] small fixes in documentation | drebs | |
2017-10-17 | [ci] remove duplicated call to responsiveness tests | drebs | |
2017-10-17 | [ci] improve outlier detection output | drebs | |
2017-10-16 | [ci] fix benchmarks jobs, remove wrong mark filtering | drebs | |
2017-10-16 | [bug] use all default server config values | drebs | |
Server config dictionary was being poorly updated, and not all default values were being added in runtime. This was mainly a problem in tests, but fixing may avoid possible bugs with this implementation in the future. | |||
2017-10-16 | [ci] compare benchmark results with history | drebs | |
2017-10-12 | [test] remove now unneeded semaphored from benchmarks | drebs | |
With the introduction of semaphores in the blobmanager level, there's no need for them in the benchmark tests now. | |||
2017-10-12 | [bug] limit number of concurrent requests to local db | drebs | |
When running stress tests on blobs storage, we get weird errors when hundreds of requests are made concurrently to the sqlite backend. This commit adds a limit so only 10 requests will be delivered to the backend at a time. | |||
2017-10-12 | [feature] make concurrent blob writes configurable | drebs | |
2017-10-11 | [bug] limit concurrent blob writes in server | drebs | |
If there's no limit to the number of concurrent blob writes in the server, the maximum limit of open files will eventually be reached, and the processing of requests will start crashing. This commit adds a semaphore to limit the number of concurrent writes in the server. Related: #8973 | |||
2017-10-11 | [ci] avoid leaking secrets | drebs | |
With the set -o xtrace option all the variables are printed to stdout, thus leaking all secrets in the CI output. This commit removes that option from relevant scripts. | |||
2017-10-11 | [doc] install deps from wheels when building docs | drebs | |
If pyzqm is built with --install-option, then all dependencies will be forced to install from source. By removing that option, we allow pip to use wheels. | |||
2017-10-11 | [doc] add script to build docs | drebs | |
2017-10-10 | [test] add docstring for blobs tests | drebs | |
2017-10-10 | [test] fix blobs download benchmark test | drebs | |
2017-10-10 | [feature] log OS errors when writing blobs | drebs | |
2017-10-10 | [bug] handle put errors in the incoming blobs api | drebs | |
An errback was missing in the PUT renderer method of the incoming API. Because of that, requests to that endpoint were not being correctly finished in case of errors when writing blobs. That was causing delivery requests to hang until timeout. Closes: #8977 | |||
2017-10-09 | [ci] spit commands to stdout on ci shell scripts | drebs | |
2017-10-09 | [pkg] add timestamps to packaging jobs | drebs | |
2017-10-09 | [test] only run benchmark stage on scheduled pipelines | drebs | |
2017-10-09 | [doc] remove variable name from storage secret JSON example | drebs | |
2017-10-09 | [test] rename blobs sqlite benchmarks file | drebs | |