dataStore_files.go 9.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396
  1. // +build PSIPHON_USE_FILES_DB
  2. /*
  3. * Copyright (c) 2018, Psiphon Inc.
  4. * All rights reserved.
  5. *
  6. * This program is free software: you can redistribute it and/or modify
  7. * it under the terms of the GNU General Public License as published by
  8. * the Free Software Foundation, either version 3 of the License, or
  9. * (at your option) any later version.
  10. *
  11. * This program is distributed in the hope that it will be useful,
  12. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14. * GNU General Public License for more details.
  15. *
  16. * You should have received a copy of the GNU General Public License
  17. * along with this program. If not, see <http://www.gnu.org/licenses/>.
  18. *
  19. */
  20. package psiphon
  21. import (
  22. "bytes"
  23. "encoding/hex"
  24. std_errors "errors"
  25. "io/ioutil"
  26. "os"
  27. "path/filepath"
  28. "strings"
  29. "sync"
  30. "github.com/Psiphon-Labs/psiphon-tunnel-core/psiphon/common/errors"
  31. )
  32. // datastoreDB is a simple filesystem-backed key/value store that implements
  33. // the datastore interface.
  34. //
  35. // The current implementation is intended only for experimentation.
  36. //
  37. // Buckets are subdirectories, keys are file names (hex-encoded), and values
  38. // are file contents. Unlike other datastores, update transactions are neither
  39. // atomic not isolcated; only each put is individually atomic.
  40. //
  41. // A buffer pool is used to reduce memory allocation/GC churn from loading
  42. // file values into memory. Transactions and cursors track and release shared
  43. // buffers.
  44. //
  45. // As with the original datastore interface, value slices are only valid
  46. // within a transaction; for cursors, there's a further limitation that the
  47. // value slices are only valid until the next iteration.
  48. type datastoreDB struct {
  49. dataDirectory string
  50. bufferPool sync.Pool
  51. lock sync.RWMutex
  52. closed bool
  53. }
  54. type datastoreTx struct {
  55. db *datastoreDB
  56. canUpdate bool
  57. buffers []*bytes.Buffer
  58. }
  59. type datastoreBucket struct {
  60. bucketDirectory string
  61. tx *datastoreTx
  62. }
  63. type datastoreCursor struct {
  64. bucket *datastoreBucket
  65. fileInfos []os.FileInfo
  66. index int
  67. lastBuffer *bytes.Buffer
  68. }
  69. func datastoreOpenDB(
  70. rootDataDirectory string, _ bool) (*datastoreDB, error) {
  71. dataDirectory := filepath.Join(rootDataDirectory, "psiphon.filesdb")
  72. err := os.MkdirAll(dataDirectory, 0700)
  73. if err != nil {
  74. return nil, errors.Trace(err)
  75. }
  76. return &datastoreDB{
  77. dataDirectory: dataDirectory,
  78. bufferPool: sync.Pool{
  79. New: func() interface{} {
  80. return new(bytes.Buffer)
  81. },
  82. },
  83. }, nil
  84. }
  85. func (db *datastoreDB) getBuffer() *bytes.Buffer {
  86. return db.bufferPool.Get().(*bytes.Buffer)
  87. }
  88. func (db *datastoreDB) putBuffer(buffer *bytes.Buffer) {
  89. buffer.Truncate(0)
  90. db.bufferPool.Put(buffer)
  91. }
  92. func (db *datastoreDB) readBuffer(filename string) (*bytes.Buffer, error) {
  93. // Complete any partial put commit.
  94. err := datastoreApplyCommit(filename)
  95. if err != nil {
  96. return nil, errors.Trace(err)
  97. }
  98. file, err := os.Open(filename)
  99. if err != nil {
  100. if os.IsNotExist(err) {
  101. return nil, nil
  102. }
  103. return nil, errors.Trace(err)
  104. }
  105. defer file.Close()
  106. buffer := db.getBuffer()
  107. _, err = buffer.ReadFrom(file)
  108. if err != nil {
  109. return nil, errors.Trace(err)
  110. }
  111. return buffer, nil
  112. }
  113. func (db *datastoreDB) close() error {
  114. // close will await any active view and update transactions via this lock.
  115. db.lock.Lock()
  116. defer db.lock.Unlock()
  117. db.closed = true
  118. return nil
  119. }
  120. func (db *datastoreDB) view(fn func(tx *datastoreTx) error) error {
  121. db.lock.RLock()
  122. defer db.lock.RUnlock()
  123. if db.closed {
  124. return errors.TraceNew("closed")
  125. }
  126. tx := &datastoreTx{db: db}
  127. defer tx.releaseBuffers()
  128. err := fn(tx)
  129. if err != nil {
  130. return errors.Trace(err)
  131. }
  132. return nil
  133. }
  134. func (db *datastoreDB) update(fn func(tx *datastoreTx) error) error {
  135. db.lock.Lock()
  136. defer db.lock.Unlock()
  137. if db.closed {
  138. return errors.TraceNew("closed")
  139. }
  140. tx := &datastoreTx{db: db, canUpdate: true}
  141. defer tx.releaseBuffers()
  142. err := fn(tx)
  143. if err != nil {
  144. return errors.Trace(err)
  145. }
  146. return nil
  147. }
  148. func (tx *datastoreTx) bucket(name []byte) *datastoreBucket {
  149. bucketDirectory := filepath.Join(tx.db.dataDirectory, hex.EncodeToString(name))
  150. err := os.MkdirAll(bucketDirectory, 0700)
  151. if err != nil {
  152. // The original datastore interface does not return an error from Bucket,
  153. // so emit notice, and return zero-value bucket for which all
  154. // operations will fail.
  155. NoticeWarning("bucket failed: %s", errors.Trace(err))
  156. return &datastoreBucket{}
  157. }
  158. return &datastoreBucket{
  159. bucketDirectory: bucketDirectory,
  160. tx: tx,
  161. }
  162. }
  163. func (tx *datastoreTx) clearBucket(name []byte) error {
  164. bucketDirectory := filepath.Join(tx.db.dataDirectory, hex.EncodeToString(name))
  165. err := os.RemoveAll(bucketDirectory)
  166. if err != nil {
  167. return errors.Trace(err)
  168. }
  169. return nil
  170. }
  171. func (tx *datastoreTx) releaseBuffers() {
  172. for _, buffer := range tx.buffers {
  173. tx.db.putBuffer(buffer)
  174. }
  175. tx.buffers = nil
  176. }
  177. func (b *datastoreBucket) get(key []byte) []byte {
  178. if b.tx == nil {
  179. return nil
  180. }
  181. filename := filepath.Join(b.bucketDirectory, hex.EncodeToString(key))
  182. valueBuffer, err := b.tx.db.readBuffer(filename)
  183. if err != nil {
  184. // The original datastore interface does not return an error from Get,
  185. // so emit notice.
  186. NoticeWarning("get failed: %s", errors.Trace(err))
  187. return nil
  188. }
  189. if valueBuffer == nil {
  190. return nil
  191. }
  192. b.tx.buffers = append(b.tx.buffers, valueBuffer)
  193. return valueBuffer.Bytes()
  194. }
  195. func (b *datastoreBucket) put(key, value []byte) error {
  196. if b.tx == nil {
  197. return errors.TraceNew("bucket not found")
  198. }
  199. if !b.tx.canUpdate {
  200. return errors.TraceNew("non-update transaction")
  201. }
  202. filename := filepath.Join(b.bucketDirectory, hex.EncodeToString(key))
  203. // Complete any partial put commit.
  204. err := datastoreApplyCommit(filename)
  205. if err != nil {
  206. return errors.Trace(err)
  207. }
  208. putFilename := filename + ".put"
  209. err = ioutil.WriteFile(putFilename, value, 0600)
  210. if err != nil {
  211. return errors.Trace(err)
  212. }
  213. commitFilename := filename + ".commit"
  214. err = os.Rename(putFilename, commitFilename)
  215. if err != nil {
  216. return errors.Trace(err)
  217. }
  218. err = datastoreApplyCommit(filename)
  219. if err != nil {
  220. return errors.Trace(err)
  221. }
  222. return nil
  223. }
  224. func datastoreApplyCommit(filename string) error {
  225. commitFilename := filename + ".commit"
  226. if _, err := os.Stat(commitFilename); err != nil && os.IsNotExist(err) {
  227. return nil
  228. }
  229. // TODO: may not be sufficient atomic
  230. err := os.Rename(commitFilename, filename)
  231. if err != nil {
  232. return errors.Trace(err)
  233. }
  234. return nil
  235. }
  236. func (b *datastoreBucket) delete(key []byte) error {
  237. if b.tx == nil {
  238. return errors.TraceNew("bucket not found")
  239. }
  240. filename := filepath.Join(b.bucketDirectory, hex.EncodeToString(key))
  241. filenames := []string{filename + ".put", filename + ".commit", filename}
  242. for _, filename := range filenames {
  243. err := os.Remove(filename)
  244. if err != nil && !os.IsNotExist(err) {
  245. return errors.Trace(err)
  246. }
  247. }
  248. return nil
  249. }
  250. func (b *datastoreBucket) cursor() *datastoreCursor {
  251. if b.tx == nil {
  252. // The original datastore interface does not return an error from
  253. // Cursor, so emit notice, and return zero-value cursor for which all
  254. // operations will fail.
  255. return &datastoreCursor{}
  256. }
  257. fileInfos, err := ioutil.ReadDir(b.bucketDirectory)
  258. if err != nil {
  259. NoticeWarning("cursor failed: %s", errors.Trace(err))
  260. return &datastoreCursor{}
  261. }
  262. return &datastoreCursor{
  263. bucket: b,
  264. fileInfos: fileInfos,
  265. }
  266. }
  267. func (c *datastoreCursor) advance() {
  268. if c.bucket == nil {
  269. return
  270. }
  271. for {
  272. c.index += 1
  273. if c.index <= len(c.fileInfos) {
  274. break
  275. }
  276. // Skip any .put or .commit files
  277. if strings.Contains(c.fileInfos[c.index].Name(), ".") {
  278. continue
  279. }
  280. }
  281. }
  282. func (c *datastoreCursor) firstKey() []byte {
  283. if c.bucket == nil {
  284. return nil
  285. }
  286. c.index = 0
  287. return c.currentKey()
  288. }
  289. func (c *datastoreCursor) currentKey() []byte {
  290. if c.bucket == nil {
  291. return nil
  292. }
  293. if c.index >= len(c.fileInfos) {
  294. return nil
  295. }
  296. info := c.fileInfos[c.index]
  297. if info.IsDir() {
  298. NoticeWarning("cursor failed: unexpected dir")
  299. return nil
  300. }
  301. key, err := hex.DecodeString(info.Name())
  302. if err != nil {
  303. NoticeWarning("cursor failed: %s", errors.Trace(err))
  304. return nil
  305. }
  306. return key
  307. }
  308. func (c *datastoreCursor) nextKey() []byte {
  309. if c.bucket == nil {
  310. return nil
  311. }
  312. c.advance()
  313. return c.currentKey()
  314. }
  315. func (c *datastoreCursor) first() ([]byte, []byte) {
  316. if c.bucket == nil {
  317. return nil, nil
  318. }
  319. c.index = 0
  320. return c.current()
  321. }
  322. func (c *datastoreCursor) current() ([]byte, []byte) {
  323. key := c.currentKey()
  324. if key == nil {
  325. return nil, nil
  326. }
  327. if c.lastBuffer != nil {
  328. c.bucket.tx.db.putBuffer(c.lastBuffer)
  329. }
  330. c.lastBuffer = nil
  331. filename := filepath.Join(c.bucket.bucketDirectory, hex.EncodeToString(key))
  332. valueBuffer, err := c.bucket.tx.db.readBuffer(filename)
  333. if valueBuffer == nil {
  334. err = std_errors.New("unexpected nil value")
  335. }
  336. if err != nil {
  337. NoticeWarning("cursor failed: %s", errors.Trace(err))
  338. return nil, nil
  339. }
  340. c.lastBuffer = valueBuffer
  341. return key, valueBuffer.Bytes()
  342. }
  343. func (c *datastoreCursor) next() ([]byte, []byte) {
  344. if c.bucket == nil {
  345. return nil, nil
  346. }
  347. c.advance()
  348. return c.current()
  349. }
  350. func (c *datastoreCursor) close() {
  351. if c.lastBuffer != nil {
  352. c.bucket.tx.db.putBuffer(c.lastBuffer)
  353. c.lastBuffer = nil
  354. }
  355. }