diff --git a/Dockerfile b/Dockerfile
index 6f2fcb2..086f94d 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,5 +1,7 @@
-FROM scrapinghub/base:12.04
+FROM scrapinghub/scrapinghub-stack-scrapy:1.4
 WORKDIR /app
-ADD requirements.txt /app/requirements.txt
+COPY requirements.txt /app/requirements.txt
 RUN pip install -r /app/requirements.txt
-ADD . /app
+ENV SCRAPY_SETTINGS_MODULE testspiders.settings
+COPY . /app
+RUN python setup.py install
diff --git a/requirements.txt b/requirements.txt
index 7903385..9b3b13f 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,3 +1 @@
 Scrapy>=1.0.2
-service_identity==1.0.0
-https://github.com/scrapinghub/scrapinghub-entrypoint-scrapy/tarball/7e9190e#egg=sh_scrapy
diff --git a/testspiders/spiders/broken_link.py b/testspiders/spiders/broken_link.py
new file mode 100644
index 0000000..29bddf3
--- /dev/null
+++ b/testspiders/spiders/broken_link.py
@@ -0,0 +1,79 @@
+# coding: utf8
+
+import scrapy
+
+try:
+    from urllib.parse import urlparse
+except ImportError:
+    from urlparse import urlparse
+
+
+class BrokenLink(scrapy.Spider):
+    """
+    Spider arguments:
+    - input_url: Where to start the crawl with.
+    - allowed_domains (optional): Comma-separated list of domains to restrict the crawl with. If not specified, it would be inferred from the input URL, e.g. http://doc.scrapy.org/en/latest/intro/overview.html -> doc.scrapy.org
+
+    Settings:
+    - DEPTH_LIMIT: Controls the maximum depth (defaults to 50).
+    - MAX_REQUESTS: Controls the maximum requests (defaults to 100000). The actual number of requests may be slightly different, e.g. MAX_REQUESTS=1000 and the spider stops when having sent 1008 requests.
+    """
+    name = 'broken_link'
+    custom_settings = {
+        'HTTPERROR_ALLOW_ALL': True,
+        'DEPTH_LIMIT': 50,
+        'MAX_REQUESTS': 100000,
+        'RETRY_HTTP_CODES': [],
+    }
+
+    def __init__(self, input_url, allowed_domains=None, *args, **kwargs):
+        """Initializes the instance"""
+        super(BrokenLink, self).__init__(*args, **kwargs)
+        self.start_urls = [input_url]
+        if allowed_domains:
+            self.allowed_domains = allowed_domains.split(',')
+        else:
+            netloc = urlparse(input_url).netloc
+            domain = netloc.split('@')[-1].split(':')[0]
+            self.allowed_domains = [domain]
+
+    def start_requests(self):
+        """Generates initial requests"""
+        for url in self.start_urls:
+            # Explicitly set the errback handler
+            yield scrapy.Request(
+                url,
+                dont_filter=True,
+                callback=self.parse,
+                errback=self.errback
+            )
+
+    def parse(self, response):
+        """Parses a default response"""
+        if not isinstance(response, scrapy.http.TextResponse):
+            self.crawler.stats.inc_value('non_text_response')
+            return
+        if response.status >= 400 and response.status <= 599:
+            yield {
+                'url': response.url,
+                'status': 'invalid_http_status',
+                'http_status': response.status,
+            }
+        max_reqs = self.settings.getint('MAX_REQUESTS', 0)
+        stats = self.crawler.stats
+        for href in response.css('a::attr(href)').extract():
+            if max_reqs and max_reqs < stats.get_value('scheduler/enqueued'):
+                break
+            yield scrapy.Request(
+                response.urljoin(href),
+                callback=self.parse,
+                errback=self.errback
+            )
+
+    def errback(self, err):
+        """Handles an error"""
+        return {
+            'url': err.request.url,
+            'status': 'error_downloading_http_response',
+            'message': str(err.value),
+        }
diff --git a/testspiders/spiders/followall.py b/testspiders/spiders/followall.py
index 76e5bfe..da5e4c9 100644
--- a/testspiders/spiders/followall.py
+++ b/testspiders/spiders/followall.py
@@ -63,7 +63,7 @@ def _set_title(self, page, response):
 
     def _set_new_cookies(self, page, response):
         cookies = []
-        for cookie in [x.split(';', 1)[0] for x in
+        for cookie in [x.split(b';', 1)[0] for x in
                        response.headers.getlist('Set-Cookie')]:
             if cookie not in self.cookies_seen:
                 self.cookies_seen.add(cookie)
diff --git a/testspiders/spiders/timed.py b/testspiders/spiders/timed.py
index 4396352..d18914b 100644
--- a/testspiders/spiders/timed.py
+++ b/testspiders/spiders/timed.py
@@ -2,7 +2,6 @@
 Crawll-all spider without domain restriction
 """
 from testspiders.spiders.followall import FollowAllSpider
-from twisted.internet import reactor
 
 
 class TimedSpider(FollowAllSpider):
@@ -15,6 +14,7 @@ def __init__(self, **kw):
         super(TimedSpider, self).__init__(**kw)
 
     def start_requests(self):
+        from twisted.internet import reactor
         reactor.callLater(self.timeout, self.stop)
         return super(TimedSpider, self).start_requests()
 
diff --git a/testspiders/spiders/timewaste.py b/testspiders/spiders/timewaste.py
new file mode 100644
index 0000000..6c14b0b
--- /dev/null
+++ b/testspiders/spiders/timewaste.py
@@ -0,0 +1,20 @@
+import scrapy
+
+
+class Spider(scrapy.Spider):
+    name = 'timewaste'
+    start_urls = ('https://example.com',)
+
+    def __init__(self, **kw):
+        self.timeout = int(kw.pop('timeout', '600'))
+        super(Spider, self).__init__(**kw)
+
+    def parse(self, response):
+        from twisted.internet import reactor, defer
+        self.log('I will waste your time for {} seconds'.format(self.timeout))
+        dfd = defer.Deferred()
+        reactor.callLater(self.timeout, dfd.callback, None)
+        return dfd
+
+    def stop(self):
+        self.crawler.engine.close_spider(self, 'timeout')