diff --git a/scripts/.gitignore b/scripts/.gitignore
index 8c7b543b8d..9becffee7d 100644
--- a/scripts/.gitignore
+++ b/scripts/.gitignore
@@ -1,2 +1,2 @@
-logs/
-database/
\ No newline at end of file
+/logs/
+/database/
\ No newline at end of file
diff --git a/scripts/commands/database/cleanup.js b/scripts/commands/database/cleanup.js
new file mode 100644
index 0000000000..ede11d701e
--- /dev/null
+++ b/scripts/commands/database/cleanup.js
@@ -0,0 +1,25 @@
+const { db, logger } = require('../../core')
+
+async function main() {
+  logger.info(`loading streams...`)
+  await db.streams.load()
+  let streams = await db.streams.find({})
+
+  logger.info(`removing broken links...`)
+  let removed = 0
+  const buffer = {}
+  for (const stream of streams) {
+    const duplicate = buffer[stream.channel_id]
+    if (duplicate && ['offline', 'timeout'].includes(stream.status.code)) {
+      await db.streams.remove({ _id: stream._id })
+      removed++
+    } else {
+      buffer[stream.channel_id] = stream
+    }
+  }
+  db.streams.compact()
+
+  logger.info(`removed ${removed} links`)
+}
+
+main()
diff --git a/scripts/commands/database/create.js b/scripts/commands/database/create.js
new file mode 100644
index 0000000000..7611872adf
--- /dev/null
+++ b/scripts/commands/database/create.js
@@ -0,0 +1,82 @@
+const { db, file, parser, store, logger, id } = require('../../core')
+const { program } = require('commander')
+const _ = require('lodash')
+
+const options = program
+  .option(
+    '--max-clusters <max-clusters>',
+    'Set maximum number of clusters',
+    parser.parseNumber,
+    256
+  )
+  .option('--input-dir <input-dir>', 'Set path to input directory', 'channels')
+  .parse(process.argv)
+  .opts()
+
+async function main() {
+  logger.info('starting...')
+  logger.info(`number of clusters: ${options.maxClusters}`)
+
+  await saveToDatabase(await findStreams())
+
+  logger.info('done')
+}
+
+main()
+
+async function findStreams() {
+  logger.info(`looking for streams...`)
+
+  await db.streams.load()
+  const files = await file.list(`${options.inputDir}/**/*.m3u`)
+  const streams = []
+  for (const filepath of files) {
+    const items = await parser.parsePlaylist(filepath)
+    for (const item of items) {
+      item.filepath = filepath
+      streams.push(item)
+    }
+  }
+  logger.info(`found ${streams.length} streams`)
+
+  return streams
+}
+
+async function saveToDatabase(streams = []) {
+  logger.info('saving to the database...')
+
+  await db.streams.reset()
+  const chunks = split(_.shuffle(streams), options.maxClusters)
+  for (const [i, chunk] of chunks.entries()) {
+    for (const item of chunk) {
+      const stream = store.create()
+      stream.set('channel_id', { channel_id: item.tvg.id })
+      stream.set('channel_name', { title: item.name })
+      stream.set('filepath', { filepath: item.filepath })
+      stream.set('resolution', { title: item.name })
+      stream.set('status', { title: item.name })
+      stream.set('url', { url: item.url })
+      stream.set('http', { http: item.http })
+      stream.set('is_broken', { status: stream.get('status') })
+      stream.set('updated', { updated: false })
+      stream.set('cluster_id', { cluster_id: i + 1 })
+
+      if (!stream.get('channel_id')) {
+        const channel_id = id.generate(item.name, item.filepath)
+
+        stream.set('channel_id', { channel_id })
+        stream.set('updated', { updated: true })
+      }
+
+      await db.streams.insert(stream.data())
+    }
+  }
+}
+
+function split(arr, n) {
+  let result = []
+  for (let i = n; i > 0; i--) {
+    result.push(arr.splice(0, Math.ceil(arr.length / i)))
+  }
+  return result
+}
diff --git a/scripts/commands/database/matrix.js b/scripts/commands/database/matrix.js
new file mode 100644
index 0000000000..f51e37f0b1
--- /dev/null
+++ b/scripts/commands/database/matrix.js
@@ -0,0 +1,16 @@
+const { logger, db } = require('../../core')
+
+async function main() {
+  await db.streams.load()
+  const docs = await db.streams.find({}).sort({ cluster_id: 1 })
+  const cluster_id = docs.reduce((acc, curr) => {
+    if (!acc.includes(curr.cluster_id)) acc.push(curr.cluster_id)
+    return acc
+  }, [])
+
+  const matrix = { cluster_id }
+  const output = `::set-output name=matrix::${JSON.stringify(matrix)}`
+  logger.info(output)
+}
+
+main()
diff --git a/scripts/commands/database/update.js b/scripts/commands/database/update.js
new file mode 100644
index 0000000000..c1786bc9ad
--- /dev/null
+++ b/scripts/commands/database/update.js
@@ -0,0 +1,176 @@
+const { db, store, parser, file, logger } = require('../../core')
+const statuses = require('../../data/statuses')
+const _ = require('lodash')
+
+const items = []
+
+const LOGS_DIR = process.env.LOGS_DIR || 'scripts/logs/cluster/load'
+
+async function main() {
+  let streams = await loadStreams()
+  const results = await loadResults()
+  const origins = await findOrigins(results)
+  streams = await updateStreams(streams, results, origins)
+
+  await updateDatabase(streams)
+}
+
+main()
+
+async function loadStreams() {
+  logger.info('loading streams...')
+
+  await db.streams.load()
+  const streams = await db.streams.find({})
+
+  logger.info(`found ${streams.length} streams`)
+
+  return streams
+}
+
+async function loadResults() {
+  logger.info('loading results from logs...')
+
+  const results = {}
+  const files = await file.list(`${LOGS_DIR}/cluster_*.log`)
+  for (const filepath of files) {
+    const parsed = await parser.parseLogs(filepath)
+    for (const item of parsed) {
+      results[item._id] = item
+    }
+  }
+
+  logger.info(`found ${Object.values(results).length} results`)
+
+  return results
+}
+
+async function findOrigins(results = {}) {
+  logger.info('searching for stream origins...')
+
+  const origins = {}
+  for (const { error, requests } of Object.values(results)) {
+    if (error || !Array.isArray(requests) || !requests.length) continue
+
+    let origin = requests.shift()
+    origin = new URL(origin.url)
+    for (const request of requests) {
+      const curr = new URL(request.url)
+      const key = curr.href.replace(/(^\w+:|^)/, '')
+      if (!origins[key] && curr.host === origin.host) {
+        origins[key] = origin.href
+      }
+    }
+  }
+
+  logger.info(`found ${_.uniq(Object.values(origins)).length} origins`)
+
+  return origins
+}
+
+async function updateStreams(items = [], results = {}, origins = {}) {
+  logger.info('updating streams...')
+
+  let updated = 0
+  const output = []
+  for (const item of items) {
+    const stream = store.create(item)
+    const result = results[item._id]
+
+    if (result) {
+      const { error, streams, requests } = result
+      const resolution = parseResolution(streams)
+      const origin = findOrigin(requests, origins)
+      let status = parseStatus(error)
+
+      if (status) {
+        const prevStatus = item.status
+        if (prevStatus.code === 'not_247')
+          // not_247 -> * = not_247
+          status = item.status
+        else if (prevStatus.code === 'geo_blocked')
+          // geo_blocked -> * = geo_blocked
+          status = item.status
+        else if (status.code === 'geo_blocked')
+          // * -> geo_blocked = *
+          status = item.status
+        else if (prevStatus.code === 'offline' && status.code === 'online')
+          // offline -> online = not_247
+          status = statuses['not_247']
+
+        stream.set('status', { status })
+        stream.set('is_broken', { status: stream.get('status') })
+      }
+
+      if (resolution) {
+        stream.set('resolution', { resolution })
+      }
+
+      if (origin) {
+        stream.set('url', { url: origin })
+      }
+    }
+
+    if (stream.changed) {
+      stream.set('updated', true)
+      output.push(stream.data())
+      updated++
+    }
+  }
+
+  logger.info(`updated ${updated} streams`)
+
+  return output
+}
+
+async function updateDatabase(streams = []) {
+  logger.info('updating database...')
+
+  for (const stream of streams) {
+    await db.streams.update({ _id: stream._id }, stream)
+  }
+  db.streams.compact()
+
+  logger.info('done')
+}
+
+function findOrigin(requests = [], origins = {}) {
+  if (origins && Array.isArray(requests)) {
+    requests = requests.map(r => r.url.replace(/(^\w+:|^)/, ''))
+    for (const url of requests) {
+      if (origins[url]) {
+        return origins[url]
+      }
+    }
+  }
+
+  return null
+}
+
+function parseResolution(streams) {
+  const resolution = streams
+    .filter(s => s.codec_type === 'video')
+    .reduce(
+      (acc, curr) => {
+        if (curr.height > acc.height) return { width: curr.width, height: curr.height }
+        return acc
+      },
+      { width: 0, height: 0 }
+    )
+
+  if (resolution.width > 0 && resolution.height > 0) return resolution
+  return null
+}
+
+function parseStatus(error) {
+  if (error) {
+    if (error.includes('timed out')) {
+      return statuses['timeout']
+    } else if (error.includes('403')) {
+      return statuses['geo_blocked']
+    }
+    return statuses['offline']
+  }
+
+  return statuses['online']
+}