http_global_client_util.go 13 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553
  1. package http
  2. import (
  3. "compress/gzip"
  4. "context"
  5. "encoding/json"
  6. "errors"
  7. "fmt"
  8. "sync"
  9. "github.com/seaweedfs/seaweedfs/weed/util"
  10. "github.com/seaweedfs/seaweedfs/weed/util/mem"
  11. "github.com/seaweedfs/seaweedfs/weed/util/request_id"
  12. "io"
  13. "net/http"
  14. "net/url"
  15. "strings"
  16. "time"
  17. "github.com/seaweedfs/seaweedfs/weed/glog"
  18. "github.com/seaweedfs/seaweedfs/weed/security"
  19. )
  20. var ErrNotFound = fmt.Errorf("not found")
  21. var (
  22. jwtSigningReadKey security.SigningKey
  23. jwtSigningReadKeyExpires int
  24. loadJwtConfigOnce sync.Once
  25. )
  26. func loadJwtConfig() {
  27. v := util.GetViper()
  28. jwtSigningReadKey = security.SigningKey(v.GetString("jwt.signing.read.key"))
  29. jwtSigningReadKeyExpires = v.GetInt("jwt.signing.read.expires_after_seconds")
  30. }
  31. func Post(url string, values url.Values) ([]byte, error) {
  32. r, err := GetGlobalHttpClient().PostForm(url, values)
  33. if err != nil {
  34. return nil, err
  35. }
  36. defer r.Body.Close()
  37. b, err := io.ReadAll(r.Body)
  38. if r.StatusCode >= 400 {
  39. if err != nil {
  40. return nil, fmt.Errorf("%s: %d - %s", url, r.StatusCode, string(b))
  41. } else {
  42. return nil, fmt.Errorf("%s: %s", url, r.Status)
  43. }
  44. }
  45. if err != nil {
  46. return nil, err
  47. }
  48. return b, nil
  49. }
  50. // github.com/seaweedfs/seaweedfs/unmaintained/repeated_vacuum/repeated_vacuum.go
  51. // may need increasing http.Client.Timeout
  52. func Get(url string) ([]byte, bool, error) {
  53. return GetAuthenticated(url, "")
  54. }
  55. func GetAuthenticated(url, jwt string) ([]byte, bool, error) {
  56. request, err := http.NewRequest(http.MethodGet, url, nil)
  57. if err != nil {
  58. return nil, true, err
  59. }
  60. maybeAddAuth(request, jwt)
  61. request.Header.Add("Accept-Encoding", "gzip")
  62. response, err := GetGlobalHttpClient().Do(request)
  63. if err != nil {
  64. return nil, true, err
  65. }
  66. defer CloseResponse(response)
  67. var reader io.ReadCloser
  68. switch response.Header.Get("Content-Encoding") {
  69. case "gzip":
  70. reader, err = gzip.NewReader(response.Body)
  71. if err != nil {
  72. return nil, true, err
  73. }
  74. defer reader.Close()
  75. default:
  76. reader = response.Body
  77. }
  78. b, err := io.ReadAll(reader)
  79. if response.StatusCode >= 400 {
  80. retryable := response.StatusCode >= 500
  81. return nil, retryable, fmt.Errorf("%s: %s", url, response.Status)
  82. }
  83. if err != nil {
  84. return nil, false, err
  85. }
  86. return b, false, nil
  87. }
  88. func Head(url string) (http.Header, error) {
  89. r, err := GetGlobalHttpClient().Head(url)
  90. if err != nil {
  91. return nil, err
  92. }
  93. defer CloseResponse(r)
  94. if r.StatusCode >= 400 {
  95. return nil, fmt.Errorf("%s: %s", url, r.Status)
  96. }
  97. return r.Header, nil
  98. }
  99. func maybeAddAuth(req *http.Request, jwt string) {
  100. if jwt != "" {
  101. req.Header.Set("Authorization", "BEARER "+string(jwt))
  102. }
  103. }
  104. func Delete(url string, jwt string) error {
  105. req, err := http.NewRequest(http.MethodDelete, url, nil)
  106. maybeAddAuth(req, jwt)
  107. if err != nil {
  108. return err
  109. }
  110. resp, e := GetGlobalHttpClient().Do(req)
  111. if e != nil {
  112. return e
  113. }
  114. defer resp.Body.Close()
  115. body, err := io.ReadAll(resp.Body)
  116. if err != nil {
  117. return err
  118. }
  119. switch resp.StatusCode {
  120. case http.StatusNotFound, http.StatusAccepted, http.StatusOK:
  121. return nil
  122. }
  123. m := make(map[string]interface{})
  124. if e := json.Unmarshal(body, &m); e == nil {
  125. if s, ok := m["error"].(string); ok {
  126. return errors.New(s)
  127. }
  128. }
  129. return errors.New(string(body))
  130. }
  131. func DeleteProxied(url string, jwt string) (body []byte, httpStatus int, err error) {
  132. req, err := http.NewRequest(http.MethodDelete, url, nil)
  133. maybeAddAuth(req, jwt)
  134. if err != nil {
  135. return
  136. }
  137. resp, err := GetGlobalHttpClient().Do(req)
  138. if err != nil {
  139. return
  140. }
  141. defer resp.Body.Close()
  142. body, err = io.ReadAll(resp.Body)
  143. if err != nil {
  144. return
  145. }
  146. httpStatus = resp.StatusCode
  147. return
  148. }
  149. func GetBufferStream(url string, values url.Values, allocatedBytes []byte, eachBuffer func([]byte)) error {
  150. r, err := GetGlobalHttpClient().PostForm(url, values)
  151. if err != nil {
  152. return err
  153. }
  154. defer CloseResponse(r)
  155. if r.StatusCode != 200 {
  156. return fmt.Errorf("%s: %s", url, r.Status)
  157. }
  158. for {
  159. n, err := r.Body.Read(allocatedBytes)
  160. if n > 0 {
  161. eachBuffer(allocatedBytes[:n])
  162. }
  163. if err != nil {
  164. if err == io.EOF {
  165. return nil
  166. }
  167. return err
  168. }
  169. }
  170. }
  171. func GetUrlStream(url string, values url.Values, readFn func(io.Reader) error) error {
  172. r, err := GetGlobalHttpClient().PostForm(url, values)
  173. if err != nil {
  174. return err
  175. }
  176. defer CloseResponse(r)
  177. if r.StatusCode != 200 {
  178. return fmt.Errorf("%s: %s", url, r.Status)
  179. }
  180. return readFn(r.Body)
  181. }
  182. func DownloadFile(fileUrl string, jwt string) (filename string, header http.Header, resp *http.Response, e error) {
  183. req, err := http.NewRequest(http.MethodGet, fileUrl, nil)
  184. if err != nil {
  185. return "", nil, nil, err
  186. }
  187. maybeAddAuth(req, jwt)
  188. response, err := GetGlobalHttpClient().Do(req)
  189. if err != nil {
  190. return "", nil, nil, err
  191. }
  192. header = response.Header
  193. contentDisposition := response.Header["Content-Disposition"]
  194. if len(contentDisposition) > 0 {
  195. idx := strings.Index(contentDisposition[0], "filename=")
  196. if idx != -1 {
  197. filename = contentDisposition[0][idx+len("filename="):]
  198. filename = strings.Trim(filename, "\"")
  199. }
  200. }
  201. resp = response
  202. return
  203. }
  204. func Do(req *http.Request) (resp *http.Response, err error) {
  205. return GetGlobalHttpClient().Do(req)
  206. }
  207. func NormalizeUrl(url string) (string, error) {
  208. return GetGlobalHttpClient().NormalizeHttpScheme(url)
  209. }
  210. func ReadUrl(ctx context.Context, fileUrl string, cipherKey []byte, isContentCompressed bool, isFullChunk bool, offset int64, size int, buf []byte) (int64, error) {
  211. if cipherKey != nil {
  212. var n int
  213. _, err := readEncryptedUrl(ctx, fileUrl, "", cipherKey, isContentCompressed, isFullChunk, offset, size, func(data []byte) {
  214. n = copy(buf, data)
  215. })
  216. return int64(n), err
  217. }
  218. req, err := http.NewRequest(http.MethodGet, fileUrl, nil)
  219. if err != nil {
  220. return 0, err
  221. }
  222. if !isFullChunk {
  223. req.Header.Add("Range", fmt.Sprintf("bytes=%d-%d", offset, offset+int64(size)-1))
  224. } else {
  225. req.Header.Set("Accept-Encoding", "gzip")
  226. }
  227. r, err := GetGlobalHttpClient().Do(req)
  228. if err != nil {
  229. return 0, err
  230. }
  231. defer CloseResponse(r)
  232. if r.StatusCode >= 400 {
  233. return 0, fmt.Errorf("%s: %s", fileUrl, r.Status)
  234. }
  235. var reader io.ReadCloser
  236. contentEncoding := r.Header.Get("Content-Encoding")
  237. switch contentEncoding {
  238. case "gzip":
  239. reader, err = gzip.NewReader(r.Body)
  240. if err != nil {
  241. return 0, err
  242. }
  243. defer reader.Close()
  244. default:
  245. reader = r.Body
  246. }
  247. var (
  248. i, m int
  249. n int64
  250. )
  251. // refers to https://github.com/golang/go/blob/master/src/bytes/buffer.go#L199
  252. // commit id c170b14c2c1cfb2fd853a37add92a82fd6eb4318
  253. for {
  254. m, err = reader.Read(buf[i:])
  255. i += m
  256. n += int64(m)
  257. if err == io.EOF {
  258. return n, nil
  259. }
  260. if err != nil {
  261. return n, err
  262. }
  263. if n == int64(len(buf)) {
  264. break
  265. }
  266. }
  267. // drains the response body to avoid memory leak
  268. data, _ := io.ReadAll(reader)
  269. if len(data) != 0 {
  270. glog.V(1).InfofCtx(ctx, "%s reader has remaining %d bytes", contentEncoding, len(data))
  271. }
  272. return n, err
  273. }
  274. func ReadUrlAsStream(ctx context.Context, fileUrl string, cipherKey []byte, isContentGzipped bool, isFullChunk bool, offset int64, size int, fn func(data []byte)) (retryable bool, err error) {
  275. return ReadUrlAsStreamAuthenticated(ctx, fileUrl, "", cipherKey, isContentGzipped, isFullChunk, offset, size, fn)
  276. }
  277. func ReadUrlAsStreamAuthenticated(ctx context.Context, fileUrl, jwt string, cipherKey []byte, isContentGzipped bool, isFullChunk bool, offset int64, size int, fn func(data []byte)) (retryable bool, err error) {
  278. if cipherKey != nil {
  279. return readEncryptedUrl(ctx, fileUrl, jwt, cipherKey, isContentGzipped, isFullChunk, offset, size, fn)
  280. }
  281. req, err := http.NewRequest(http.MethodGet, fileUrl, nil)
  282. maybeAddAuth(req, jwt)
  283. if err != nil {
  284. return false, err
  285. }
  286. if isFullChunk {
  287. req.Header.Add("Accept-Encoding", "gzip")
  288. } else {
  289. req.Header.Add("Range", fmt.Sprintf("bytes=%d-%d", offset, offset+int64(size)-1))
  290. }
  291. request_id.InjectToRequest(ctx, req)
  292. r, err := GetGlobalHttpClient().Do(req)
  293. if err != nil {
  294. return true, err
  295. }
  296. defer CloseResponse(r)
  297. if r.StatusCode >= 400 {
  298. if r.StatusCode == http.StatusNotFound {
  299. return true, fmt.Errorf("%s: %s: %w", fileUrl, r.Status, ErrNotFound)
  300. }
  301. retryable = r.StatusCode >= 499
  302. return retryable, fmt.Errorf("%s: %s", fileUrl, r.Status)
  303. }
  304. var reader io.ReadCloser
  305. contentEncoding := r.Header.Get("Content-Encoding")
  306. switch contentEncoding {
  307. case "gzip":
  308. reader, err = gzip.NewReader(r.Body)
  309. defer reader.Close()
  310. default:
  311. reader = r.Body
  312. }
  313. var (
  314. m int
  315. )
  316. buf := mem.Allocate(64 * 1024)
  317. defer mem.Free(buf)
  318. for {
  319. // Check for context cancellation before each read
  320. select {
  321. case <-ctx.Done():
  322. return false, ctx.Err()
  323. default:
  324. }
  325. m, err = reader.Read(buf)
  326. if m > 0 {
  327. fn(buf[:m])
  328. }
  329. if err == io.EOF {
  330. return false, nil
  331. }
  332. if err != nil {
  333. return true, err
  334. }
  335. }
  336. }
  337. func readEncryptedUrl(ctx context.Context, fileUrl, jwt string, cipherKey []byte, isContentCompressed bool, isFullChunk bool, offset int64, size int, fn func(data []byte)) (bool, error) {
  338. encryptedData, retryable, err := GetAuthenticated(fileUrl, jwt)
  339. if err != nil {
  340. return retryable, fmt.Errorf("fetch %s: %v", fileUrl, err)
  341. }
  342. decryptedData, err := util.Decrypt(encryptedData, util.CipherKey(cipherKey))
  343. if err != nil {
  344. return false, fmt.Errorf("decrypt %s: %v", fileUrl, err)
  345. }
  346. if isContentCompressed {
  347. decryptedData, err = util.DecompressData(decryptedData)
  348. if err != nil {
  349. glog.V(0).InfofCtx(ctx, "unzip decrypt %s: %v", fileUrl, err)
  350. }
  351. }
  352. if len(decryptedData) < int(offset)+size {
  353. return false, fmt.Errorf("read decrypted %s size %d [%d, %d)", fileUrl, len(decryptedData), offset, int(offset)+size)
  354. }
  355. if isFullChunk {
  356. fn(decryptedData)
  357. } else {
  358. sliceEnd := int(offset) + size
  359. fn(decryptedData[int(offset):sliceEnd])
  360. }
  361. return false, nil
  362. }
  363. func ReadUrlAsReaderCloser(fileUrl string, jwt string, rangeHeader string) (*http.Response, io.ReadCloser, error) {
  364. req, err := http.NewRequest(http.MethodGet, fileUrl, nil)
  365. if err != nil {
  366. return nil, nil, err
  367. }
  368. if rangeHeader != "" {
  369. req.Header.Add("Range", rangeHeader)
  370. } else {
  371. req.Header.Add("Accept-Encoding", "gzip")
  372. }
  373. maybeAddAuth(req, jwt)
  374. r, err := GetGlobalHttpClient().Do(req)
  375. if err != nil {
  376. return nil, nil, err
  377. }
  378. if r.StatusCode >= 400 {
  379. CloseResponse(r)
  380. return nil, nil, fmt.Errorf("%s: %s", fileUrl, r.Status)
  381. }
  382. var reader io.ReadCloser
  383. contentEncoding := r.Header.Get("Content-Encoding")
  384. switch contentEncoding {
  385. case "gzip":
  386. reader, err = gzip.NewReader(r.Body)
  387. if err != nil {
  388. return nil, nil, err
  389. }
  390. default:
  391. reader = r.Body
  392. }
  393. return r, reader, nil
  394. }
  395. func CloseResponse(resp *http.Response) {
  396. if resp == nil || resp.Body == nil {
  397. return
  398. }
  399. reader := &CountingReader{reader: resp.Body}
  400. io.Copy(io.Discard, reader)
  401. resp.Body.Close()
  402. if reader.BytesRead > 0 {
  403. glog.V(1).Infof("response leftover %d bytes", reader.BytesRead)
  404. }
  405. }
  406. func CloseRequest(req *http.Request) {
  407. reader := &CountingReader{reader: req.Body}
  408. io.Copy(io.Discard, reader)
  409. req.Body.Close()
  410. if reader.BytesRead > 0 {
  411. glog.V(1).Infof("request leftover %d bytes", reader.BytesRead)
  412. }
  413. }
  414. type CountingReader struct {
  415. reader io.Reader
  416. BytesRead int
  417. }
  418. func (r *CountingReader) Read(p []byte) (n int, err error) {
  419. n, err = r.reader.Read(p)
  420. r.BytesRead += n
  421. return n, err
  422. }
  423. func RetriedFetchChunkData(ctx context.Context, buffer []byte, urlStrings []string, cipherKey []byte, isGzipped bool, isFullChunk bool, offset int64, fileId string) (n int, err error) {
  424. loadJwtConfigOnce.Do(loadJwtConfig)
  425. var jwt security.EncodedJwt
  426. if len(jwtSigningReadKey) > 0 {
  427. jwt = security.GenJwtForVolumeServer(
  428. jwtSigningReadKey,
  429. jwtSigningReadKeyExpires,
  430. fileId,
  431. )
  432. }
  433. var shouldRetry bool
  434. for waitTime := time.Second; waitTime < util.RetryWaitTime; waitTime += waitTime / 2 {
  435. // Check for context cancellation before starting retry loop
  436. select {
  437. case <-ctx.Done():
  438. return n, ctx.Err()
  439. default:
  440. }
  441. for _, urlString := range urlStrings {
  442. // Check for context cancellation before each volume server request
  443. select {
  444. case <-ctx.Done():
  445. return n, ctx.Err()
  446. default:
  447. }
  448. n = 0
  449. if strings.Contains(urlString, "%") {
  450. urlString = url.PathEscape(urlString)
  451. }
  452. shouldRetry, err = ReadUrlAsStreamAuthenticated(ctx, urlString+"?readDeleted=true", string(jwt), cipherKey, isGzipped, isFullChunk, offset, len(buffer), func(data []byte) {
  453. // Check for context cancellation during data processing
  454. select {
  455. case <-ctx.Done():
  456. // Stop processing data when context is cancelled
  457. return
  458. default:
  459. }
  460. if n < len(buffer) {
  461. x := copy(buffer[n:], data)
  462. n += x
  463. }
  464. })
  465. if !shouldRetry {
  466. break
  467. }
  468. if err != nil {
  469. glog.V(0).InfofCtx(ctx, "read %s failed, err: %v", urlString, err)
  470. } else {
  471. break
  472. }
  473. }
  474. if err != nil && shouldRetry {
  475. glog.V(0).InfofCtx(ctx, "retry reading in %v", waitTime)
  476. // Sleep with proper context cancellation and timer cleanup
  477. timer := time.NewTimer(waitTime)
  478. select {
  479. case <-ctx.Done():
  480. timer.Stop()
  481. return n, ctx.Err()
  482. case <-timer.C:
  483. // Continue with retry
  484. }
  485. } else {
  486. break
  487. }
  488. }
  489. return n, err
  490. }