Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
50 most recent check-ins
|
2020-08-07
| ||
| 22:04 | initial empty check-in Leaf check-in: 739a24e10c user: adizaimi tags: trunk | |
| 21:03 | Make testrunner.c compile (ansi strictness), remove python script (#2394) * testrunner.c in some systems fails to compile due to ansi strictness, so fixing Makefile and .c for that. * Updating the tests/Readme how to use the testrunner to run tests. * Deleted obsolete python script tests/tools/testrunner Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> Leaf check-in: a0a15a34a0 user: azaimi@bloomberg.net tags: trunk, origin/az_cleanup_testrunner, mponomar/cmake, origin/AZ_makecluster | |
| 17:05 | Abort when we cant convert from client to server type (#2389) All types should have a conversion from client to server. In the case that we add new types, we should add conversion as well, so we should abort here to catch cases who forget to add conversion. Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> Leaf check-in: bf77ace5c1 user: azaimi@bloomberg.net tags: trunk, origin/az_test_abort_client2server_type, mponomar/cmake, origin/AZ_makecluster | |
| 16:13 | Log in statreq write count for every table (#2224) * Use statthd thread to update saved_write_count and aa_new_count * Show count diffs in statreq file * Cleanup that logic from autoanalyze thread, update test * Use atomic ops to update write_count for more accurate accounting * Refactor function to print diff to statreq and update prev value, dont print DB number * Get schema_lk in statthd when walking table list * Refactor printing of OSQL type via array of strings instead of case stmt The new write counts will appear as such: ``` 04/30 13:20:57: DIFF REQUEST STATS FOR TABLE 't1' 04/30 13:20:57: inserted rows 32 04/30 13:20:57: updated rows 32 04/30 13:20:57: deleted rows 32 ``` Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> Leaf check-in: 55894fdbe6 user: azaimi@bloomberg.net tags: trunk, az_write_count_statthd, mponomar/cmake, origin/AZ_makecluster | |
| 15:52 | These instructions also give a successful "make test" on Ubuntu 20.04 and Debian 10. Signed-off-by: danshearer <dan@shearer.org> check-in: 53b18800c3 user: azaimi@bloomberg.net tags: trunk, origin/az_test_abort_client2server_type, mponomar/cmake, origin/AZ_makecluster | |
| 01:30 | use macro to expand enum and string representation of types Leaf check-in: a1c21fadbc user: azaimi@bloomberg.net tags: trunk, az_osql_rpl_types, mponomar/cmake, origin/AZ_makecluster | |
|
2020-08-06
| ||
| 20:32 | Merge pull request #2388 from mistachkin/num_wrk_thds Keep track of the number of thread-pool threads that are actively performing work items. check-in: 76994a33bf user: noreply@github.com tags: trunk, origin/az_test_abort_client2server_type, mponomar/cmake, origin/AZ_makecluster | |
| 17:16 | Add comdb2_functions table and autocomplete ability for them Add new system table `comdb2_functions` table. Add ability to autocomplete all sqlite and comdb2 functions from cdb2sql command line. Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> Leaf check-in: 4eece0c7e6 user: azaimi@bloomberg.net tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 14:57 | Merge remote-tracking branch 'upstream/master' into az_dbstore_function Leaf check-in: 4e5b48d934 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 14:56 | disable this part again check-in: 9ccc4fde47 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 13:56 | cmake: FindSQLite3 for cmake version >= 3.14 Signed-off-by: Akshat Sikarwar <asikarwar1@bloomberg.net> check-in: 3d8d2b85cd user: asikarwar1@bloomberg.net tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
|
2020-08-05
| ||
| 19:51 | Keep track of the number of thread-pool threads that are actively performing work items. Signed-off-by: Joe Mistachkin <joe@mistachkin.com> check-in: 6ede0a20ce user: joe@mistachkin.com tags: trunk, origin/az_test_abort_client2server_type, mponomar/cmake, origin/AZ_makecluster | |
| 18:22 | cmake: Use FindSQLite3 Signed-off-by: Akshat Sikarwar <asikarwar1@bloomberg.net> check-in: f4f763a316 user: asikarwar1@bloomberg.net tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 18:22 | cmake: Remove tinfo Signed-off-by: Akshat Sikarwar <asikarwar1@bloomberg.net> check-in: 4cefb78407 user: asikarwar1@bloomberg.net tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 00:39 | change so we can pass to sqlite anything between {}, but alas no checks check-in: fe15871392 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
|
2020-08-04
| ||
| 18:43 | trying to pass string parameters to functions for dbstore check-in: d90d66bbed user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 14:33 | Merge pull request #2012 from mistachkin/sc_queues_commit_fix Prevent access to freed memory after trans_commit / trans_abort check-in: 32c584ca6d user: noreply@github.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 13:59 | put gbl as it was check-in: 77a370ab2d user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 13:51 | fix doc and a check check-in: 6a30908c2e user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 12:53 | Since the trans_commit() pipeline (eventually) frees its 'tran' argument, make sure to NULL it out so it will not be passed to trans_abort(). Signed-off-by: Joe Mistachkin <joe@mistachkin.com> check-in: 0277238985 user: joe@mistachkin.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 02:04 | disable this part for now check-in: 92a6d48b4a user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
|
2020-08-03
| ||
| 23:00 | db not reachable after adding tbl, sleep 1 seems to be needed for now check-in: b251ad0e8c user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 17:42 | Allocator index is 1-based Fixes #2246 Signed-off-by: Rivers Zhang <hzhang320@bloomberg.net> Leaf check-in: f9af55efb4 user: hzhang320@bloomberg.net tags: trunk, upstream/7.0, mponomar/cmake, origin/AZ_makecluster | |
| 17:41 | Allocator index is 1-based Fixes #2246 Signed-off-by: Rivers Zhang <hzhang320@bloomberg.net> check-in: 5063cc3a5f user: riverszhang89@users.noreply.github.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 17:41 | Merge remote-tracking branch 'upstream/master' into az_dbstore_function check-in: 21136f5b2a user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 15:46 | Optionally verify record count (#2362) Make record count verification optional to temporarily suppress our recnums index corruption tickets. (DRQS 144891579) Signed-off-by: Rivers Zhang <hzhang320@bloomberg.net> check-in: 5905a81a4b user: hzhang320@bloomberg.net tags: trunk, upstream/7.0, mponomar/cmake, origin/AZ_makecluster | |
| 14:30 | more cleanup of spaces and not needed code check-in: 0a17f3ee2b user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 13:31 | some cleanup check-in: c07e51b626 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
| 03:51 | got guid() working from csc2 by changing how its handled in .l file check-in: fe36a953bf user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
|
2020-08-02
| ||
| 14:35 | need to have dbstore={guid{}} to not have syntax error with next token null=no check-in: 12e3ec49b4 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
|
2020-07-31
| ||
| 20:15 | Reporting size of temporary files (#2270) Two approaches: 1. via system table ```sql $ cdb2sql mydb 'SELECT * FROM comdb2_temporary_file_sizes' (type='temptables', bytes=0) (type='sqlsorters', bytes=0) (type='blkseqs', bytes=131072) (type='others', bytes=0) ``` 2. via stat size ``` $ send mydb stat size table sqlite_stat1 sz 72.00KB 0% (dta 64.00KB, ix0 8.00KB) table sqlite_stat4 sz 24.00KB 0% (dta 16.00KB, ix0 8.00KB) table t sz 8.00KB 0% (dta 8.00KB) table t1 sz 8.00KB 0% (dta 8.00KB) temp files sz 128.00KB 0% (tmptbls 0B, sqlsorters 0B, blkseqs 128.00KB, others 0B) log files sz 15.58MB 98% (1 logs) GRAND TOTAL 15.81MB FILESYSTEM SIZE 29.40GB, AVAILABLE SPACE 13.46GB ``` (DRQS 158892990) Signed-off-by: Rivers Zhang <hzhang320@bloomberg.net> check-in: 8e0a3a84af user: riverszhang89@users.noreply.github.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 17:30 | Workaround for MacOS + CMake + TINFO Signed-off-by: Michael Ponomarenko <mponomarenko@bloomberg.net> check-in: 60f189b5d0 user: mponomar@users.noreply.github.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 17:19 | Merge pull request #2378 from mistachkin/nextseq_sqlite_integration_cleanup Minor cleanups to SQLite integration related to the new 'NEXTSEQUENCE' feature. check-in: 4e38b3a9d0 user: noreply@github.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 15:59 | Minor cleanups to SQLite integration related to the new 'NEXTSEQUENCE' feature. Signed-off-by: Joe Mistachkin <joe@mistachkin.com> check-in: d15fd3e9bc user: joe@mistachkin.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 14:46 | Simple sql hooks Signed-off-by: Mark Hannum <mhannum72@gmail.com> Made headway on SQLite logic Commit while I backport a fix to 7.0 Mini-demo now works Changes from review with Joe Don't overuse flags-bits Comment Make sequence a specially handled type on the server side Update logic working llmeta functions clang master columns functions utilize llmeta, deadlock handling First stab at schemachange Working cleanup a bit Start nextsequence test Test verifies basic functionality Add prevent-modify tests Prevent adding next nextsequence column which doesnt allow nulls Modify dbstore value on insert as well as update, complete fastinit test clang Fix delete table crash Fix comdb2sys and auth tests remove unneeded code Change from Rivers formatting changes Coerce ENOMEM errors in bdb_lite_fetch_keys_int to DBNOTFOUND Add a property to comdb2_columns Signed-off-by: Michael Ponomarenko <mponomarenko@bloomberg.net> Disable small sequences by default, better error messages, other review changes Fix comdb2sys test Signed-off-by: Michael Ponomarenko <mponomarenko@bloomberg.net> fix ddl_no_csc2 clang Simple changes to fix tagged api clang Incorporate feedback from Joe and Akshat WIP Return correct error message for select nextsequence Review comments from Nirbhay, Akshat, Rivers, and others Print explicit error messages (thanks Nirbhay), checkin to work on docs Regenerated bubble diagrams Fix comdb2sys and auth tests Changes from Nirbhay Update nextsequence in docs fix documentation clang Fix ddl_no_csc2 test check-in: 1d98b34cfb user: mhannum72@gmail.com tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 07:57 | saving work -- dbstore={guid()}, dbstore={randomblob(16)} work check-in: f3ac68f5ee user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_function, mponomar/cmake, origin/AZ_makecluster | |
|
2020-07-30
| ||
| 18:00 | Small cleanup in dyns_get_table_field_option function (#2375) In this cleanup: - Refactor function to be more readable - Use pointer instead of repeat structure-member-dereferencing train - Reuse length of string instead of recomputing it Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> check-in: 64517f3a46 user: azaimi@bloomberg.net tags: trunk, origin/az_comdb2_functions, mponomar/cmake, origin/AZ_makecluster | |
| 06:22 | put test back Leaf check-in: 385f747505 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |
| 06:14 | rename uuid back to guid in setting dbstore check-in: 3ba195ce0e user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |
| 05:58 | got it working by renaming sqlite function from uuid to something else check-in: 12c9de3279 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |
| 04:34 | Cleanup some more warnings Leaf check-in: 2e3d218f25 user: azaimi@bloomberg.net tags: trunk, cleanup, mponomar/cmake, origin/AZ_makecluster | |
| 04:26 | Makefile with Debug as default version to test separately Leaf check-in: de0af14387 user: azaimi@bloomberg.net tags: trunk, origin/Debug, mponomar/cmake, origin/AZ_makecluster | |
| 04:22 | Eventlog: print client argv0 if it is set (#2369) If client application argv0 is set, print it in eventlog. Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> check-in: 71661fa758 user: azaimi@bloomberg.net tags: trunk, upstream/7.0, mponomar/cmake, origin/AZ_makecluster | |
| 04:19 | Eventlog: print client argv0 if it is set (#2369) If client application argv0 is set, print it in eventlog. Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> Leaf check-in: 014ab81599 user: azaimi@bloomberg.net tags: trunk, origin/az_evntlog_print_client, mponomar/cmake, origin/AZ_makecluster | |
| 01:50 | test that have not broken schema members called uuid check-in: ebe64ac202 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |
| 01:38 | Have dbstore=uuid only--no quotes, no guid check-in: aa70971bb2 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |
| 01:01 | Small cleanup in dyns_get_table_field_option function (#2375) In this cleanup: - Refactor function to be more readable - Use pointer instead of repeating variable access train - Reuse length of string instead of recomputing it Signed-off-by: Adi Zaimi <azaimi@bloomberg.net> Leaf check-in: 0d11481018 user: azaimi@bloomberg.net tags: trunk, origin/az_cleanup, mponomar/cmake, origin/AZ_makecluster | |
|
2020-07-29
| ||
| 20:12 | cdb2api bug fixes: pmux, garbage config and the local tier (#2367) The change fixes 3 bugs in cdb2api. 1. Unnecessary pmux lookup in DIRECT-CPU mode In DIRECT-CPU mode the API always looks up the database port. This costs an unnecessary round trip to pmux when there is already a cached connection in the sockpool. We should do so only when the API needs to make a connection on its own. 2. Garbage configuration file The API attempts to process "(null).cfg". This is caused by passing a NULL database name to snprintf when formatting the database configuration file path. 3. The "local" tier does not mean localhost The API performs dbinfo query and pulls all nodes in the cluster regardless of tier. However dbinfo should be skipped when "local" is specified as tier. Signed-off-by: Rivers Zhang <hzhang320@bloomberg.net> check-in: 5529a686c4 user: riverszhang89@users.noreply.github.com tags: trunk, origin/az_evntlog_print_client, mponomar/cmake, origin/AZ_makecluster | |
| 18:36 | fix default guid primary key issue with extra spaces after guid check-in: 1a79764927 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |
| 17:59 | case insensitive guid/uuid, add primary key to tests so they will fail on dup check-in: 6a753d56e9 user: azaimi@bloomberg.net tags: trunk, origin/az_dbstore_guid, mponomar/cmake, origin/AZ_makecluster | |