Commit Graph
100 Commits
Author SHA1 Message Date
Vitaliy Filippov c79b38bd26 Move all sources to subdirs 2024-05-15 11:06:01 +03:00
Vitaliy Filippov 44692d148a Make vitastor_kv.h header public 2024-05-15 01:49:38 +03:00
Vitaliy Filippov ba52359611 Fix last master commit 2024-05-15 01:49:31 +03:00
Vitaliy Filippov 23a9aa93b5 Fix pool create/modify --block_size validation 2024-05-04 16:33:22 +03:00
Vitaliy Filippov 2412d9e239 Fix TTL comparison for lease/keepalive 2024-04-30 01:53:05 +03:00
Vitaliy Filippov 9301c857b1 Release 1.6.1
A bunch of monitor fixes

- Add noout flag for OSDs (/vitastor/config/osd/xx)
- Fix "effective" size of degraded PGs (and thus "used space") calculation in monitor
- Fix monitor not clearing PGs of deleted pools
- Fix incorrect PG generation with hosts with 0 OSDs
- Fix monitor crashing during primary OSD recheck when pool has no PGs
- Fix monitor crashing when node_placement included non-existing OSDs
- Fix possible data movement after removing OSDs reweighted to 0
- Remove extra empty keys from pool configurations created by vitastor-cli create-pool
- Fix 32-bit build
2024-04-22 02:01:29 +03:00
Vitaliy Filippov 3094358ec2 Fix autovivification leading to extra empty keys in pool-create 2024-04-20 02:04:09 +03:00
Vitaliy Filippov 87f666d2a2 Filter out OSDs reweighted to 0 2024-04-20 02:03:53 +03:00
Vitaliy Filippov bd7fe4ef8f Filter out non-existing OSDs added in node_placement 2024-04-20 02:03:36 +03:00
Vitaliy Filippov 1b3f9a1416 Do not set non-existing OSD weight to 0, we'll remove them instead 2024-04-20 02:03:11 +03:00
Vitaliy Filippov a7b7354f38 Do not recheck primary distribution when pool has no PGs 2024-04-20 02:02:47 +03:00
Vitaliy Filippov 765befa22f Remove empty nodes from tree because PG DSL expects that all leaf nodes are OSDs 2024-04-20 02:02:28 +03:00
Vitaliy Filippov 87b3ab94fe Do not disable require-atomic-updates and no-unused-vars 2024-04-20 02:02:13 +03:00
Vitaliy Filippov 2c0801f6e4 Configure ESLint and add it to CI 2024-04-16 02:39:31 +03:00
Vitaliy Filippov fd83fef1d9 Fix pool deletion 2024-04-16 02:20:26 +03:00
Vitaliy Filippov 8d1067971b Fix pg_effsize (and thus "used space") calculation in monitor 2024-04-16 02:20:18 +03:00
Vitaliy Filippov ae5af04fde Add noout flag for OSDs 2024-04-16 02:19:55 +03:00
Vitaliy Filippov 266d038b11 Fix 32-bit build warnings and one error again :-) 2024-04-11 22:49:33 +03:00
Vitaliy Filippov ff4414d37e Release 1.6.0
New features:

- Implement "hierarchical failure domains" and other complex distribution rules, for example
  EC 4+2 over 3 DC, with 2 chunks per each DC ([documentation](docs/config/pool.en.md#level_placement))
- Make OSDs handle ENOSPC - now cluster stays online even if some OSDs fill up
  to 100 %, only writes requiring free space hang
- Implement Stage/Unstage & volume locking for CSI to prevent parallel mounting
  and/or modifications of the same volume
- Warn about full and almost full OSDs in vitastor-cli status
- Add an experimental NBD netlink map mode as an option ([documentation](docs/usage/nbd.en.md))
- Add --pg parameter to vitastor-cli describe, print objects with 0x in human-readable format too
- Add [administration docs](docs/usage/admin.en.md)

Bug fixes:

- Fix client operation retry timeout - previously the timeout wasn't applied and writes were
  retries almost instantly
- Fix monitors crashing on invalid pool configurations
- Fix journaling - make each journal write wait for all previous journal writes
- Fix monitor thinking that OSD weight is 0 after deleting /osd/config/ key online
- Fix a write stall caused by flusher possibly not trimming journal on rollback
- Set 32k csum_block_size for HDD by default in vitastor-disk
2024-04-09 16:57:59 +03:00
Vitaliy Filippov 0fa7ecc03f Add also a test for OSD tags 2024-04-09 16:57:59 +03:00
Vitaliy Filippov c29bfe12eb Oops - fix filter_by_root_node, add a test for it 2024-04-09 15:48:44 +03:00
Vitaliy Filippov 57bf84ddb2 Fix filtering in mon 2024-04-09 14:51:05 +03:00
Vitaliy Filippov dff4879c8c Check if NBD_ATTR_BACKEND_IDENTIFIER is defined 2024-04-09 13:16:58 +03:00
Vitaliy Filippov af9a853db6 Move NBD netlink map&unmap to separate commands, add "netlink-revive" command 2024-04-08 16:34:41 +03:00
Vitaliy Filippov b7a3275af3 Make netlink optional 2024-04-08 01:51:28 +03:00
Vitaliy Filippov 64c5c4ca26 Fix code style 2024-04-08 01:35:03 +03:00
Vitaliy Filippov 6366972fe8 Warn about full and almost full OSDs in status 2024-04-07 19:39:51 +03:00
Vitaliy Filippov 2b863fb715 Add ENOSPC handling tests 2024-04-07 19:39:33 +03:00
Vitaliy Filippov 3bf4dd5abd Fix client op retry timeout - do not retry immediately 2024-04-07 19:08:36 +03:00
Vitaliy Filippov 3b84dcaedd Handle ENOSPC during write - rollback partial EC writes, remember partial replica writes 2024-04-07 18:02:05 +03:00
Vitaliy Filippov 20fbc4a745 Add --pg parameter to vitastor-cli describe, print objects with 0x in human-readable format too 2024-04-07 12:39:46 +03:00
Vitaliy Filippov 02993ee1dd Implement Stage/Unstage & volume locking for CSI to prevent parallel modifications of the same volume 2024-04-07 11:48:19 +03:00
Vitaliy Filippov 3629dbc54d Plug the new PG combinator into monitor 2024-04-07 02:44:17 +03:00
Vitaliy Filippov 29284bef40 Implement new DSL/rule-based PG generation algorithm 2024-04-07 00:36:20 +03:00
Vitaliy Filippov 6a924d6066 Extract PG combinator into a separate module 2024-04-07 00:36:20 +03:00
Vitaliy Filippov 9fe779a691 Do not die on invalid pool configurations 2024-04-07 00:36:20 +03:00
Vitaliy Filippov 31c2751b9b Move NBD/VDUSE map/unmap functions to a separate file 2024-04-07 00:36:09 +03:00
Vitaliy Filippov c5195666cd Fix journal sequencing: make each journal write wait for all previous journal writes 2024-04-06 23:53:12 +03:00
Vitaliy Filippov f36d7eb76c Fix monitor thinking that OSD weight is 0 after deleting /osd/config/ key 2024-04-05 23:14:46 +03:00
Vitaliy Filippov dd7f651de1 Add --max-request-bytes=104857600 to etcd params in tests 2024-04-05 23:14:46 +03:00
Vitaliy Filippov a2994ecd0d Fix flusher possibly not trimming journal on rollback 2024-04-05 23:14:39 +03:00
Vitaliy Filippov 5d3aaf016b Add administration docs 2024-03-31 01:54:52 +03:00
Vitaliy Filippov 0b097ca3f2 Set 32k csum_block_size for HDD by default 2024-03-30 16:16:49 +03:00
Vitaliy Filippov 989675a780 s/etcd_ws_keepalive_timeout/etcd_ws_keepalive_interval/ in docs 2024-03-26 01:56:08 +03:00
Vitaliy Filippov f8c403ec9e Add newer benchmark results 2024-03-23 18:28:48 +03:00
Vitaliy Filippov bfbb85e653 Replace -Oanything with -O3, not just -O/-O1/-O2 2024-03-18 02:03:44 +03:00
Vitaliy Filippov 9ad6822353 Release 1.5.0
After half a year of hard work, VitastorFS is finally here ! :-)

New features:
- VitastorFS, a full-featured clustered (read-write-many) file system.
  Documentation: [VitastorFS](docs/usage/nfs.en.md)
- Embedded key-value database implementation based on Parallel Optimistic B-Tree
  algorithm and used for the metadata of VitastorFS
- Pool management commands in vitastor-cli (create-pool, list-pools, rm-pool, modify-pool).
  Thanks MIND Software (https://mindsw.io) for their contribution!
  [Documentation](docs/usage/cli.en.md#create-pool)

Bug fixes:
- Fix a very rare "infinite loop" in the client library
- Fix a rare OSD hang on during start when zeroing out bad metadata entries left from the previous run
2024-03-16 15:35:10 +03:00
Vitaliy Filippov 2043b4e374 Fix build errors for gcc 8 2024-03-16 15:35:10 +03:00
Vitaliy Filippov de840e6fe3 Reduce kv-cli loadjson load parallelism to 16 2024-03-16 15:35:10 +03:00
Vitaliy Filippov b5e04bf809 Fix build warning 2024-03-16 15:35:10 +03:00
Vitaliy Filippov 8807a1623b Fix markdown tables 2024-03-16 15:35:10 +03:00
Vitaliy Filippov f12855c31b Add vitastor-kv to packages 2024-03-16 15:35:10 +03:00
Vitaliy Filippov e75dcc9a71 Add documentation for VitastorFS 2024-03-16 15:16:43 +03:00
Vitaliy Filippov 88516ab4bd Remove extra log 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 6221126b4f Allow to print simple-offsets just given the device size 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 6783d4a13c Implement fool protection for FS pools 2024-03-16 13:24:36 +03:00
Vitaliy Filippov dcbe1afac3 Store pool ID in inode metadata 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 0bde28c24a Make nfs_do_rmw a library function 2024-03-16 13:24:36 +03:00
Vitaliy Filippov bb8ca6184e Support setattr guard 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 87310ef7bb Support ctime 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 4f4b2dab80 Log NFS liveness checks 2024-03-16 13:24:36 +03:00
Vitaliy Filippov f70da82317 Add loadjson command to vitastor-kv 2024-03-16 13:24:36 +03:00
Vitaliy Filippov e42148f347 Allow to specify KV commands on command line 2024-03-16 13:24:36 +03:00
Vitaliy Filippov c289584469 Add JSON dump format 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 018e89f867 Erase verf key left from creation from ientries on every modification 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 603dc68f11 Implement async mtime change 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 7b12342933 Allow to specify additional NFS mount options 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 44bf0f16ee Fix malloc/free in nfs_kv_read/write 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 8840c84572 Fix "bad key in etcd" in mon for FS pools 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 5b747c12ec Check if already mounted before mounting 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 05f5f46162 Fix zero used space, update mtime when moving/changing inode 2024-03-16 13:24:36 +03:00
Vitaliy Filippov b5604191c8 Ignore ECANCELED in nfs-proxy (happens in io_uring on fork) 2024-03-16 13:24:36 +03:00
Vitaliy Filippov e871de27de Support unaligned shared_offsets, align shared file data instead of header 2024-03-16 13:24:36 +03:00
Vitaliy Filippov f600ce98e2 Implement auto-unmount local NFS server mode for vitastor-nfs 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 57605a5c13 Return error on failed shrink 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 29bd4561bb Implement rename over an existing file/directory 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 7142460ec8 Support --logfile in nfs-proxy 2024-03-16 13:24:36 +03:00
Vitaliy Filippov d03f19ebe5 Fix shared file overlap, add FIXMEs 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 88f9d18be3 Create inode, then direntry, not direntry, then inode; retry ID collisions 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 6213fbd8c6 Fix NFS shared/aligned write FIXMEs 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 3aee37eadd Allow to disable per-inode stats for VitastorFS pools 2024-03-16 13:24:36 +03:00
Vitaliy Filippov ecfc753e93 Add basic NFS tests, fix bugs 2024-03-16 13:24:36 +03:00
Vitaliy Filippov a574f9ad71 Return block NFS implementation back as an option too 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 7c235c9103 Move KV FS header into a separate file 2024-03-16 13:24:36 +03:00
Vitaliy Filippov e5bb986164 Implement packing small files into shared inodes 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 181795d748 Split new NFS proxy implementation into multiple files 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 8cdc38805b WIP VitastorFS with metadata storage in VitastorKV 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 0cd455d17f First just recheck version without actually re-reading block in vitastor-kv 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 32ba653ba6 Fix vitastor-kv hang on reopen & unfinished closed listing 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 231d4b15fc Add loadable dump format to vitastor-kv (dump) 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 9dc4d5fd7b Fix freeing r/w buffers on errors in kv_db 2024-03-16 13:24:36 +03:00
Vitaliy Filippov e58538fa47 Fix eviction when random_pos selects the end 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 11ac9e7024 Implement min/max list_count to make listings during performance test reasonable 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 511bc3df1c Fix and improve parallel allocation
- Do not try to allocate more DB blocks in an inode block until it's "confirmed" and "locked" by the first write
- Do not recheck for new zero DB blocks on first write into an inode block - a CAS failure means someone else is already writing into it
- Throw new allocation blocks away regardless of whether the known_version is 0 on a CAS failure
2024-03-16 13:24:36 +03:00
Vitaliy Filippov a64f0d1f73 Implement key_prefix for K/V stress test 2024-03-16 13:24:36 +03:00
Vitaliy Filippov ec5f7c6b87 More fixes
- do not overwrite a block with older version if known version is newer
  (read may start before update and end after update)
- invalidated block versions can't be remembered and trusted
- right boundary for split blocks is right_half when diving down, not key_lt
- restart update also when block is "invalidated", not just on version mismatch
- copy callback in listings to avoid closure destruction bugs too
2024-03-16 13:24:36 +03:00
Vitaliy Filippov 3ebed9a749 Add logging and one more assert 2024-03-16 13:24:36 +03:00
Vitaliy Filippov eab67a6e8f Make get_block() wait for updating when unrelated block is found along the path 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 20993d9b7a Fix a race condition where changed blocks were parsed over existing cached blocks and getting a mix of data 2024-03-16 13:24:36 +03:00
Vitaliy Filippov 5cf9b343c0 Simplify code by removing an unneeded "optimisation" 2024-03-16 13:24:36 +03:00