@cloudant/couchbackup 2.9.10-SNAPSHOT.113 → 2.9.10

package/.scannerwork/scanner-report/source-6.txt

@@ -1,4 +1,4 @@
-// Copyright © 2017, 2018 IBM Corp. All rights reserved.
+// Copyright © 2017, 2021 IBM Corp. All rights reserved.
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -11,165 +11,154 @@
 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 // See the License for the specific language governing permissions and
 // limitations under the License.
-
-// Small script which backs up a Cloudant or CouchDB database to an S3
-// bucket via a stream rather than on-disk file.
-//
-// The script generates the backup object name by combining together the path
-// part of the database URL and the current time.
-
 'use strict';
 
+const async = require('async');
 const stream = require('stream');
-const url = require('url');
-
-const AWS = require('aws-sdk');
-const couchbackup = require('@cloudant/couchbackup');
-const debug = require('debug')('s3-backup');
-const VError = require('verror').VError;
-
-/*
-  Main function, run from base of file.
-*/
-function main() {
-  const argv = require('yargs')
-    .usage('Usage: $0 [options]')
-    .example('$0 -s https://user:pass@host/db -b <bucket>', 'Backup db to bucket')
-    .options({
-      source: { alias: 's', nargs: 1, demandOption: true, describe: 'Source database URL' },
-      bucket: { alias: 'b', nargs: 1, demandOption: true, describe: 'Destination bucket' },
-      prefix: { alias: 'p', nargs: 1, describe: 'Prefix for backup object key', default: 'couchbackup' },
-      s3url: { nargs: 1, describe: 'S3 endpoint URL' },
-      awsprofile: { nargs: 1, describe: 'The profile section to use in the ~/.aws/credentials file', default: 'default' }
-    })
-    .help('h').alias('h', 'help')
-    .epilog('Copyright (C) IBM 2017')
-    .argv;
-
-  const sourceUrl = argv.source;
-  const backupName = new url.URL(sourceUrl).pathname.split('/').filter(function(x) { return x; }).join('-');
-  const backupBucket = argv.bucket;
-  const backupKeyPrefix = `${argv.prefix}-${backupName}`;
-  const shallow = argv.shallow;
-
-  const backupKey = `${backupKeyPrefix}-${new Date().toISOString()}`;
-
-  const s3Endpoint = argv.s3url;
-  const awsProfile = argv.awsprofile;
-
-  // Creds are from ~/.aws/credentials, environment etc. (see S3 docs).
-  const awsOpts = {
-    signatureVersion: 'v4',
-    credentials: new AWS.SharedIniFileCredentials({ profile: awsProfile })
-  };
-  if (typeof s3Endpoint !== 'undefined') {
-    awsOpts.endpoint = new AWS.Endpoint(s3Endpoint);
-  }
-  const s3 = new AWS.S3(awsOpts);
-
-  debug(`Creating a new backup of ${s(sourceUrl)} at ${backupBucket}/${backupKey}...`);
-  bucketAccessible(s3, backupBucket)
-    .then(() => {
-      return backupToS3(sourceUrl, s3, backupBucket, backupKey, shallow);
-    })
-    .then(() => {
-      debug('done.');
-    })
-    .catch((reason) => {
-      debug(`Error: ${reason}`);
-      process.exit(1);
-    });
-}
-
-/**
- * Return a promise that resolves if the bucket is available and
- * rejects if not.
- *
- * @param {any} s3 S3 client object
- * @param {any} bucketName Bucket name
- * @returns Promise
- */
-function bucketAccessible(s3, bucketName) {
-  return new Promise(function(resolve, reject) {
-    const params = {
-      Bucket: bucketName
-    };
-    s3.headBucket(params, function(err, data) {
-      if (err) {
-        reject(new VError(err, 'S3 bucket not accessible'));
-      } else {
-        resolve();
+const error = require('./error.js');
+const debug = require('debug')('couchbackup:writer');
+
+module.exports = function(db, bufferSize, parallelism, ee) {
+  const writer = new stream.Transform({ objectMode: true });
+  let buffer = [];
+  let written = 0;
+  let linenumber = 0;
+
+  // this is the queue of chunks that are written to the database
+  // the queue's payload will be an array of documents to be written,
+  // the size of the array will be bufferSize. The variable parallelism
+  // determines how many HTTP requests will occur at any one time.
+  const q = async.queue(function(payload, cb) {
+    // if we are restoring known revisions, we need to supply new_edits=false
+    if (payload.docs && payload.docs[0] && payload.docs[0]._rev) {
+      payload.new_edits = false;
+      debug('Using new_edits false mode.');
+    }
+
+    if (!didError) {
+      db.service.postBulkDocs({
+        db: db.db,
+        bulkDocs: payload
+      }).then(response => {
+        if (!response.result || (payload.new_edits === false && response.result.length > 0)) {
+          throw new Error(`Error writing batch with new_edits:${payload.new_edits !== false}` +
+            ` and ${response.result ? response.result.length : 'unavailable'} items`);
+        }
+        written += payload.docs.length;
+        writer.emit('restored', { documents: payload.docs.length, total: written });
+        cb();
+      }).catch(err => {
+        err = error.convertResponseError(err);
+        debug(`Error writing docs ${err.name} ${err.message}`);
+        cb(err, payload);
+      });
+    }
+  }, parallelism);
+
+  let didError = false;
+
+  // write the contents of the buffer to CouchDB in blocks of bufferSize
+  function processBuffer(flush, callback) {
+    function taskCallback(err, payload) {
+      if (err && !didError) {
+        debug(`Queue task failed with error ${err.name}`);
+        didError = true;
+        q.kill();
+        writer.emit('error', err);
       }
-    });
-  });
-}
-
-/**
- * Backup directly from Cloudant to an object store object via a stream.
- *
- * @param {any} sourceUrl URL of database
- * @param {any} s3Client Object store client
- * @param {any} s3Bucket Backup destination bucket
- * @param {any} s3Key Backup destination key name (shouldn't exist)
- * @param {any} shallow Whether to use the couchbackup `shallow` mode
- * @returns Promise
- */
-function backupToS3(sourceUrl, s3Client, s3Bucket, s3Key, shallow) {
-  return new Promise((resolve, reject) => {
-    debug(`Setting up S3 upload to ${s3Bucket}/${s3Key}`);
-
-    // A pass through stream that has couchbackup's output
-    // written to it and it then read by the S3 upload client.
-    // It has a 64MB highwater mark to allow for fairly
-    // uneven network connectivity.
-    const streamToUpload = new stream.PassThrough({ highWaterMark: 67108864 });
-
-    // Set up S3 upload.
-    const params = {
-      Bucket: s3Bucket,
-      Key: s3Key,
-      Body: streamToUpload
-    };
-    s3Client.upload(params, function(err, data) {
-      debug('Object store upload done');
-      if (err) {
-        debug(err);
-        reject(new VError(err, 'Object store upload failed'));
-        return;
+    }
+
+    if (flush || buffer.length >= bufferSize) {
+      // work through the buffer to break off bufferSize chunks
+      // and feed the chunks to the queue
+      do {
+        // split the buffer into bufferSize chunks
+        const toSend = buffer.splice(0, bufferSize);
+
+        // and add the chunk to the queue
+        debug(`Adding ${toSend.length} to the write queue.`);
+        q.push({ docs: toSend }, taskCallback);
+      } while (buffer.length >= bufferSize);
+
+      // send any leftover documents to the queue
+      if (flush && buffer.length > 0) {
+        debug(`Adding remaining ${buffer.length} to the write queue.`);
+        q.push({ docs: buffer }, taskCallback);
       }
-      debug('Object store upload succeeded');
-      debug(data);
-      resolve();
-    }).httpUploadProgress = (progress) => {
-      debug(`Object store upload progress: ${progress}`);
-    };
-
-    debug(`Starting streaming data from ${s(sourceUrl)}`);
-    couchbackup.backup(
-      sourceUrl,
-      streamToUpload,
-      (err, obj) => {
-        if (err) {
-          debug(err);
-          reject(new VError(err, 'CouchBackup failed with an error'));
-          return;
+
+      // wait until the queue size falls to a reasonable level
+      async.until(
+        // wait until the queue length drops to twice the paralellism
+        // or until empty on the last write
+        function(callback) {
+          // if we encountered an error, stop this until loop
+          if (didError) {
+            return callback(null, true);
+          }
+          if (flush) {
+            callback(null, q.idle() && q.length() === 0);
+          } else {
+            callback(null, q.length() <= parallelism * 2);
+          }
+        },
+        function(cb) {
+          setTimeout(cb, 20);
+        },
+
+        function() {
+          if (flush && !didError) {
+            writer.emit('finished', { total: written });
+          }
+          // callback when we're happy with the queue size
+          callback();
+        });
+    } else {
+      callback();
+    }
+  }
+
+  // take an object
+  writer._transform = function(obj, encoding, done) {
+    // each obj that arrives here is a line from the backup file
+    // it should contain an array of objects. The length of the array
+    // depends on the bufferSize at backup time.
+    linenumber++;
+    if (!didError && obj !== '') {
+      // see if it parses as JSON
+      try {
+        const arr = JSON.parse(obj);
+
+        // if it's an array with a length
+        if (typeof arr === 'object' && arr.length > 0) {
+          // push each document into a buffer
+          buffer = buffer.concat(arr);
+
+          // pause the stream
+          // it's likely that the speed with which data can be read from disk
+          // may exceed the rate it can be written to CouchDB. To prevent
+          // the whole file being buffered in memory, we pause the stream here.
+          // it is resumed, when processBuffer calls back and we call done()
+          this.pause();
+
+          // break the buffer in to bufferSize chunks to be written to the database
+          processBuffer(false, done);
+        } else {
+          ee.emit('error', new error.BackupError('BackupFileJsonError', `Error on line ${linenumber} of backup file - not an array`));
+          done();
         }
-        debug(`Download from ${s(sourceUrl)} complete.`);
-        streamToUpload.end(); // must call end() to complete upload.
-        // resolve() is called by the upload
+      } catch (e) {
+        ee.emit('error', new error.BackupError('BackupFileJsonError', `Error on line ${linenumber} of backup file - cannot parse as JSON`));
+        // Could be an incomplete write that was subsequently resumed
+        done();
       }
-    );
-  });
-}
-
-/**
- * Remove creds from a URL, e.g., before logging
- *
- * @param {string} url URL to safen
- */
-function s(originalUrl) {
-  const parts = new url.URL(originalUrl);
-  return url.format(parts, { auth: false });
-}
-
-main();
+    } else {
+      done();
+    }
+  };
+
+  // called when we need to flush everything
+  writer._flush = function(done) {
+    processBuffer(true, done);
+  };
+  return writer;
+};

package/.scannerwork/scanner-report/source-60.txt

@@ -1,4 +1,4 @@
-// Copyright © 2017 IBM Corp. All rights reserved.
+// Copyright © 2017, 2018 IBM Corp. All rights reserved.
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -12,24 +12,164 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-/* global describe it */
+// Small script which backs up a Cloudant or CouchDB database to an S3
+// bucket via a stream rather than on-disk file.
+//
+// The script generates the backup object name by combining together the path
+// part of the database URL and the current time.
+
 'use strict';
 
-const assert = require('assert');
-const logfilegetbatches = require('../includes/logfilegetbatches.js');
-
-describe('#unit Fetching batches from a log file', function() {
-  it('should fetch multiple batches correctly', function(done) {
-    logfilegetbatches('./test/fixtures/test.log', [1, 4], function(err, data) {
-      assert.ok(!err);
-      assert.ok(data);
-      assert.strictEqual(typeof data, 'object');
-      assert.strictEqual(Object.keys(data).length, 2);
-      assert.deepStrictEqual(data['1'].docs, [{ id: '6' }, { id: '7' }, { id: '8' }, { id: '9' }, { id: '10' }]);
-      assert.strictEqual(data['1'].batch, 1);
-      assert.deepStrictEqual(data['4'].docs, [{ id: '21' }, { id: '22' }]);
-      assert.strictEqual(data['4'].batch, 4);
-      done();
+const stream = require('stream');
+const url = require('url');
+
+const AWS = require('aws-sdk');
+const couchbackup = require('@cloudant/couchbackup');
+const debug = require('debug')('s3-backup');
+const VError = require('verror').VError;
+
+/*
+  Main function, run from base of file.
+*/
+function main() {
+  const argv = require('yargs')
+    .usage('Usage: $0 [options]')
+    .example('$0 -s https://user:pass@host/db -b <bucket>', 'Backup db to bucket')
+    .options({
+      source: { alias: 's', nargs: 1, demandOption: true, describe: 'Source database URL' },
+      bucket: { alias: 'b', nargs: 1, demandOption: true, describe: 'Destination bucket' },
+      prefix: { alias: 'p', nargs: 1, describe: 'Prefix for backup object key', default: 'couchbackup' },
+      s3url: { nargs: 1, describe: 'S3 endpoint URL' },
+      awsprofile: { nargs: 1, describe: 'The profile section to use in the ~/.aws/credentials file', default: 'default' }
+    })
+    .help('h').alias('h', 'help')
+    .epilog('Copyright (C) IBM 2017')
+    .argv;
+
+  const sourceUrl = argv.source;
+  const backupName = new url.URL(sourceUrl).pathname.split('/').filter(function(x) { return x; }).join('-');
+  const backupBucket = argv.bucket;
+  const backupKeyPrefix = `${argv.prefix}-${backupName}`;
+  const shallow = argv.shallow;
+
+  const backupKey = `${backupKeyPrefix}-${new Date().toISOString()}`;
+
+  const s3Endpoint = argv.s3url;
+  const awsProfile = argv.awsprofile;
+
+  // Creds are from ~/.aws/credentials, environment etc. (see S3 docs).
+  const awsOpts = {
+    signatureVersion: 'v4',
+    credentials: new AWS.SharedIniFileCredentials({ profile: awsProfile })
+  };
+  if (typeof s3Endpoint !== 'undefined') {
+    awsOpts.endpoint = new AWS.Endpoint(s3Endpoint);
+  }
+  const s3 = new AWS.S3(awsOpts);
+
+  debug(`Creating a new backup of ${s(sourceUrl)} at ${backupBucket}/${backupKey}...`);
+  bucketAccessible(s3, backupBucket)
+    .then(() => {
+      return backupToS3(sourceUrl, s3, backupBucket, backupKey, shallow);
+    })
+    .then(() => {
+      debug('done.');
+    })
+    .catch((reason) => {
+      debug(`Error: ${reason}`);
+      process.exit(1);
+    });
+}
+
+/**
+ * Return a promise that resolves if the bucket is available and
+ * rejects if not.
+ *
+ * @param {any} s3 S3 client object
+ * @param {any} bucketName Bucket name
+ * @returns Promise
+ */
+function bucketAccessible(s3, bucketName) {
+  return new Promise(function(resolve, reject) {
+    const params = {
+      Bucket: bucketName
+    };
+    s3.headBucket(params, function(err, data) {
+      if (err) {
+        reject(new VError(err, 'S3 bucket not accessible'));
+      } else {
+        resolve();
+      }
     });
   });
-});
+}
+
+/**
+ * Backup directly from Cloudant to an object store object via a stream.
+ *
+ * @param {any} sourceUrl URL of database
+ * @param {any} s3Client Object store client
+ * @param {any} s3Bucket Backup destination bucket
+ * @param {any} s3Key Backup destination key name (shouldn't exist)
+ * @param {any} shallow Whether to use the couchbackup `shallow` mode
+ * @returns Promise
+ */
+function backupToS3(sourceUrl, s3Client, s3Bucket, s3Key, shallow) {
+  return new Promise((resolve, reject) => {
+    debug(`Setting up S3 upload to ${s3Bucket}/${s3Key}`);
+
+    // A pass through stream that has couchbackup's output
+    // written to it and it then read by the S3 upload client.
+    // It has a 64MB highwater mark to allow for fairly
+    // uneven network connectivity.
+    const streamToUpload = new stream.PassThrough({ highWaterMark: 67108864 });
+
+    // Set up S3 upload.
+    const params = {
+      Bucket: s3Bucket,
+      Key: s3Key,
+      Body: streamToUpload
+    };
+    s3Client.upload(params, function(err, data) {
+      debug('Object store upload done');
+      if (err) {
+        debug(err);
+        reject(new VError(err, 'Object store upload failed'));
+        return;
+      }
+      debug('Object store upload succeeded');
+      debug(data);
+      resolve();
+    }).httpUploadProgress = (progress) => {
+      debug(`Object store upload progress: ${progress}`);
+    };
+
+    debug(`Starting streaming data from ${s(sourceUrl)}`);
+    couchbackup.backup(
+      sourceUrl,
+      streamToUpload,
+      (err, obj) => {
+        if (err) {
+          debug(err);
+          reject(new VError(err, 'CouchBackup failed with an error'));
+          return;
+        }
+        debug(`Download from ${s(sourceUrl)} complete.`);
+        streamToUpload.end(); // must call end() to complete upload.
+        // resolve() is called by the upload
+      }
+    );
+  });
+}
+
+/**
+ * Remove creds from a URL, e.g., before logging
+ *
+ * @param {string} url URL to safen
+ */
+function s(originalUrl) {
+  const parts = new url.URL(originalUrl);
+  return url.format(parts, { auth: false });
+}
+
+main();