Commit Graph

4234 Commits

Author SHA1 Message Date
Stephen Atherton c1745b9061 Bug fix, backup snapshot dispatch shardsBehind calculation would divide by zero if the snapshot scheduled interval is 0. 2019-03-11 19:43:59 -07:00
Stephen Atherton ad4d2f192b Added release note for backup/restore datetime format changes as it breaks compatibility with existing tooling. 2019-03-11 02:10:06 -07:00
Stephen Atherton adc5553cd9 Updated backup documentation regarding timestamp format. 2019-03-11 02:05:01 -07:00
Stephen Atherton f0024c876e Changed backup paused JSON field name to be more precise and match the human readable status version. 2019-03-11 01:42:19 -07:00
Stephen Atherton f2953db7d8 Added updating of backup snapshot shards behind in snapshot dispatcher so status can determine if a snapshot is lagging the configured speed. 2019-03-11 01:25:51 -07:00
Stephen Atherton 023bbb566f Renamed backup state enums for clarity, added backup state names. Changed Epochs to EpochSeconds in backup JSON along with some other renaming/moving of fields, and added information about snapshot dispatch. Changed timestamp format for input/output in all backup/restore contexts to be a fully qualified time with timezone offset. Added information about the last snapshot dispatch to backup config and status (not yet populated). 2019-03-10 16:00:01 -07:00
Stephen Atherton 8bdb21a04e Added release note for new json output options on backup. 2019-03-06 22:44:37 -08:00
Stephen Atherton 06c11a316d Normalized timestamp to text format across backup and restore tooling. Added epochs field to JSON objects describing versions and timestamps in backup status and describe output, renamed some fields for clarity. 2019-03-06 22:34:25 -08:00
Stephen Atherton 1399aee532 Added --json option to fdbbackup status. 2019-03-06 21:32:46 -08:00
Stephen Atherton ca8bbad657 Added --json option to fdbbackup describe. Also added expired percentage indicator to snapshot details. 2019-03-06 14:14:06 -08:00
Stephen Atherton 7778112f6a Bug fix, restore was using the destination cluster to look up timestamps when printing the backup description instead of (optionally) the original cluster which generated the backup. Made missing cluster file errors more clear. 2019-03-06 02:45:55 -08:00
Stephen Atherton 888c4764d3 Bug fix: fdbrestore was not using a default range set. 2019-03-06 01:57:03 -08:00
Stephen Atherton 28e67aa041 Fixed option name to match documentation. 2019-03-05 23:26:45 -08:00
Stephen Atherton 7ffdd6c875 Added release notes for restore changes. 2019-03-05 12:55:08 -08:00
Stephen Atherton 337a5aa1fe Updated to new Database API. 2019-03-05 12:44:28 -08:00
Stephen Atherton 2f7d559442 Restore now requires an explicit destination cluster file. Restore target database version can now be specified by a timestamp, which must be resolved using the cluster file for the cluster from which the backup originated. 2019-03-05 12:28:05 -08:00
Steve Atherton 21f55e1878
Merge pull request #1190 from bnamasivayam/restore-multiple-ranges
Add support for restoring multiple ranges.
2019-03-05 10:15:55 -08:00
Evan Tschannen 69d7633d5b
Merge pull request #1217 from alexmiller-apple/tstlog-goodref
Spill-By-Reference TLog Part 4: Actually Usable Reference Spilling
2019-03-04 20:58:24 -08:00
Evan Tschannen 6d8aa20ae3
Merge pull request #1220 from bnamasivayam/dr-switchover-checks
Add some basic checks before doing an atomic switchover.
2019-03-04 20:57:50 -08:00
Evan Tschannen 827f6a81d5
Merge pull request #1230 from tclinken/lazily-fetch-health-metrics
Client lazily fetches health metrics from proxies
2019-03-04 16:46:57 -08:00
Balachandar Namasivayam 627edd4724 Add force option to 'fdbdr switch'. 2019-03-04 16:36:58 -08:00
Meng Xu afd7c1d497 AsynFileWinASIO: Make error checking consistent with Linux
In Linux, KAIO uses ASSERT to make sure open() flags have
OPEN_UNBUFFERED set.

In Windows, we uses if-condition and return io_errors() when the
flag is not set.

This PR makes Windoes implementation always use ASSERT to check the
flag.
2019-03-04 16:36:04 -08:00
Evan Tschannen 1ee9b9cda1
Merge pull request #1219 from vishesh/task/issue-1201
fix: segfault due external assignment of Endpoint::addresses #1201
2019-03-04 16:35:42 -08:00
Trevor Clinkenbeard cbc5df5e93 Removed unnecessary Trace line 2019-03-04 16:31:40 -08:00
Balachandar Namasivayam a258df32f6 Skip switchover checks for force option. 2019-03-04 15:58:36 -08:00
Vishesh Yadav 5cd8bac6cb fix: segfault due external assignment of Endpoint::addresses #1201
isLocal() now checks if the address is equal to default
NetworkAddress() which should match the behaviour before TLS changes.
2019-03-04 15:49:11 -08:00
Balachandar Namasivayam 85b18c0665
Merge pull request #1225 from bnamasivayam/master
Document behavior on the ordering of writes and read conflict ranges
2019-03-04 15:17:42 -08:00
Balachandar Namasivayam ee7ee3e783
Update documentation/sphinx/source/developer-guide.rst 2019-03-04 15:17:25 -08:00
Trevor Clinkenbeard 89cbb77b4e Merge branch 'master' of https://github.com/apple/foundationdb into lazily-fetch-health-metrics 2019-03-04 14:17:58 -08:00
Trevor Clinkenbeard 56ae46f89e Client lazily fetches health metrics from proxies 2019-03-04 14:16:39 -08:00
Evan Tschannen a6d64f2806
Merge pull request #1227 from alexmiller-apple/fix-O0-build
Fix the build on -O0.
2019-03-04 13:45:19 -08:00
Balachandar Namasivayam b0c910fba5 Update documentation. 2019-03-04 12:31:47 -08:00
Alex Miller baa3e1af2c Replace `/sizeof(Page)*sizeof(Page)` with `pageFloor()`. 2019-03-04 01:42:39 -08:00
Alex Miller ee64b43366 Change DQ shrink logic to consider "active" bytes rather than file size.
We know what the current ideal size of the DQ file should be, so we
should use it.
2019-03-04 01:42:39 -08:00
Alex Miller 244903a9de Spill txsTag by value under TagMsg/ and not TagMsgRef/
There's not a tremendous reason as to why this matters now, but I feel
like I might regret sometime later not keeping the same schema under the
same key.
2019-03-04 01:42:39 -08:00
Alex Miller 72c2cf11ab Replace ResourceLimiter with FlowLock. 2019-03-04 01:42:38 -08:00
Alex Miller 75b2546e35 Allow FlowLock::take to succeed if there's no other active requests. 2019-03-04 01:42:38 -08:00
Alex Miller 98bb58628f Upgrade FlowLock from int to int64_t. 2019-03-04 01:42:38 -08:00
Alex Miller 94bf75cb00 Allow the disk queue to shrink if it has unneeded slack space. 2019-03-04 01:42:38 -08:00
Alex Miller a462359af1 Implement g_random->randomSkewedUInt32()
This generates an exponential distribution, implemented as being
a "log-uniform" distribution.
2019-03-04 01:42:38 -08:00
Alex Miller 52d5a721a6 Don't allocate 2x the memory for a read to save 1% of allocated memory. 2019-03-04 01:42:38 -08:00
Alex Miller aff9ebe21a Spill (start,length) instead of (begin,end) to save a few bytes. 2019-03-04 01:42:38 -08:00
Alex Miller 2aa527c0ef Fix a bug resulting from concurrent TLog changes.
TLogServer was forked into OldTLogServer_6_0 at the same time that
3247d594 modified TLogServer, so the modification never made it into
OldTLogServer_6_0, resulting in a rare failure.

Manual code inspection revealed that there was also
78976161 that concurrently modified TLogServer, so that change was
copied to OldTLogServer_6_0 as well.
2019-03-04 01:42:38 -08:00
Alex Miller fb4cb8c3a8 Print out configuration changes in ConfigureTest. 2019-03-04 01:42:38 -08:00
Alex Miller 9ef283d4e7 Implement hard limiting of memory used to serve peek requests. 2019-03-04 01:42:38 -08:00
Alex Miller e3506ad9af Add a yield to parseMessagesForTag 2019-03-04 01:42:38 -08:00
Alex Miller 742f6e1847 Solve overreading via pre-calculating tag bytes per commit 2019-03-04 01:42:38 -08:00
Alex Miller e7d8520c63 Batch more when spilling data. 2019-03-04 01:42:38 -08:00
Alex Miller 71a794ccc3 Re-enable spill-by-reference testing. 2019-03-04 01:42:38 -08:00
Alex Miller 04e1170c88 Spill txsTag by value 2019-03-04 01:42:38 -08:00