How does Postges Server know to keep a database connection open - python

I wonder how does Postgres sever determine to close a DB connection, if I forgot at the Python source code side.
Does the Postgres server send a ping to the source code? From my understanding, this is not possible.

PostgreSQL indeed does something like that, although it is not a ping.
PostgreSQL uses a TCP feature called keepalive. Once enabled for a socket, the operating system kernel will regularly send keepalive messages to the other party (the peer), and if it doesn't get an answer after a couple of tries, it closes the connection.
The default timeouts for keepalive are pretty long, in the vicinity of two hours. You can configure the settings in PostgreSQL, see the documentation for details.
The default values and possible values vary according to the operating system used.
There is a similar feature available for the client side, but it is less useful and not enabled by default.

When your script quits your connection will close and the server will clean it up accordingly. Likewise, it's often the case in garbage collected languages like Python that when you stop using the connection and it falls out of scope it will be closed and cleaned up.
It is possible to write code that never releases these resources properly, that just perpetually creates new handles, something that can be problematic if you don't have something server-side that handles killing these after some period of idle time. Postgres doesn't do this by default, though it can be configured to, but MySQL does.
In short Postgres will keep a database connection open until you kill it either explicitly, such as via a close call, or implicitly, such as the handle falling out of scope and being deleted by the garbage collector.

Related

Using sniffing with python elasticsearch client to solve dead TCP connection issues

The Python elasticsearch client in my applicaiton is having connectivity issues (refused connections) because idle TCP connections timeout due to a firewall (I have no way to prevent this).
The easiest way for me to fix this would be if I could prevent the connection from going idle by sending some data over it periodically, the sniffing options in the elasticsearch client seem ideal for this, however they're not very well documented:
sniff_on_start – flag indicating whether to obtain a list of nodes
from the cluser at startup time
sniffer_timeout – number of seconds
between automatic sniffs
sniff_on_connection_fail – flag controlling
if connection failure triggers a sniff
sniff_timeout – timeout used for the sniff request - it should be a fast api call and we are talking potentially to more nodes so we want to fail quickly. Not used during initial sniffing (if sniff_on_start is on) when the connection still isn’t initialized.
What I would like is for the client to sniff every (say) 5 minutes, should I be using the sniff_timeout or sniffer_timeout option? Also, should the sniff_on_start parameter be set to True?
I used the suggestion from #val and found that these settings solved my problem:
sniff_on_start=True
sniffer_timeout=60
sniff_on_connection_fail=True
The sniffing puts enough traffic on the TCP connections so that they are never idle for long enough for our firewall to kill the conneciton.

Should a connection to Redis cluster be made on each Flask request?

I have a Flask API, it connects to a Redis cluster for caching purposes. Should I be creating and tearing down a Redis connection on each flask api call? Or, should I try and maintain a connection across requests?
My argument against the second option is that I should really try and keep the api as stateless as possible, and I also don't know if keeping some persistent across request might causes threads race conditions or other side effects.
However, if I want to persist a connection, should it be saved on the session or on the application context?
This is about performance and scale. To get those 2 buzzwords buzzing you'll in fact need persistent connections.
Eventual race conditions will be no different than with a reconnect on every request so that shouldn't be a problem. Any RCs will depend on how you're using redis, but if it's just caching there's not much room for error.
I understand the desired stateless-ness of an API from a client sides POV, but not so sure what you mean about the server side.
I'd suggest you put them in the application context, not the sessions (those could become too numerous) whereas the app context gives you the optimal 1 connection per process (and created immediately at startup). Scaling this way becomes easy-peasy: you'll never have to worry about hitting the max connection counts on the redis box (and the less multiplexing the better).
It's good idea from the performance standpoint to keep connections to a database opened between requests. The reason for that is that opening and closing connections is not free and takes some time which may become problem when you have too many requests. Another issue that a database can only handle up to a certain number of connections and if you open more, database performance will degrade, so you need to control how many connections are opened at the same time.
To solve both of these issues you may use a connection pool. A connection pool contains a number of opened database connections and provides access to them. When a database operation should be performed from a connection shoul be taken from a pool. When operation is completed a connection should be returned to the pool. If a connection is requested when all connections are taken a caller will have to wait until some connections are returned to the pool. Since no new connections are opened in this processed (they all opened in advance) this will ensure that a database will not be overloaded with too many parallel connections.
If connection pool is used correctly a single connection will be used by only one thread will use it at any moment.
Despite of the fact that connection pool has a state (it should track what connections are currently in use) your API will be stateless. This is because from the API perspective "stateless" means: does not have a state/side-effects visible to an API user. Your server can perform a number of operations that change its internal state like writing to log files or writing to a cache, but since this does not influence what data is being returned as a reply to API calls this does not make this API "stateful".
You can see some examples of using Redis connection pool here.
Regarding where it should be stored I would use application context since it fits better to its purpose.

Why do I constantly see "Resetting dropped connection" when uploading data to my database?

I'm uploading hundreds of millions of items to my database via a REST API from a cloud server on Heroku to a database in AWS EC2. I'm using Python and I am constantly seeing the following INFO log message in the logs.
[requests.packages.urllib3.connectionpool] [INFO] Resetting dropped connection: <hostname>
This "resetting of the dropped connection" seems to take many seconds (sometimes 30+ sec) before my code continues to execute again.
Firstly what exactly is happening here and why?
Secondly is there a way to stop the connection from dropping so that I am able to upload data faster?
Thanks for your help.
Andrew.
Requests uses Keep-Alive by default. Resetting dropped connection, from my understanding, means a connection that should be alive was dropped somehow. Possible reasons are:
Server doesn't support Keep-Alive.
There's no data transfer in established connections for a while, so server drops connections.
See https://stackoverflow.com/a/25239947/2142577 for more details.
The problem is really that the server has closed the connection even though the client has requested it be kept alive.
This is not necessarily because the server doesn't support keepalives, but could be that the server is configured to only allow a certain number of requests on a connection. This could be done to help spread out requests on different servers, but I think this practice is/was common as a practical defence against poorly written code that operates in the server (eg. PHP) that doesn't clean up after itself after serving a request (perhaps due to an error condition etc.)
If you think this is the case for you and you'd like to not see these logs (which are logged at INFO level), then you can add the following to quieten that part of the logging:
# Really don't need to hear about connections being brought up again after server has closed it
logging.getLogger("requests.packages.urllib3.connectionpool").setLevel(logging.WARNING)
This is common practice for services that expose RESTful APIs to avoid abuse (or DoS).
If you're stressing their API they'll drop your connection.
Try getting your script to sleep a bit every once in a while to avoid the drop.

Why does PyMongo throw AutoReconnect?

While researching some strange issues with my Python web application (in particular, issues regarding MongoDB connectivity), I noticed something on the official PyMongo documentation page. My web application uses Flask, but this shouldn't influence the issue I'm facing.
The PyMongo driver does connection pooling, but it also throws an exception (AutoReconnect) when a connection is stale and a reconnect is due.
It states that (regarding the AutoReconnect exception):
In order to auto-reconnect you must handle this exception, recognizing
that the operation which caused it has not necessarily succeeded.
Future operations will attempt to open a new connection to the
database (and will continue to raise this exception until the first
successful connection is made).
I have noticed that this actually happens constantly (and it doesn't seem to be an error). Connections are closed by the MongoDB server after what seems like several minutes of inactivity, and need to be recreated by the web application.
What I don't understand it why the PyMongo driver throws an error when it reconnects (which the user of the driver needs to handle themselves), instead of doing it transparently. (There could even be an option a user could set so that AutoReconnect exceptions do get thrown, but wouldn't a sensible default be that these exceptions don't get thrown at all, and the connections are recreated seamlessly?)
I have never encountered this behavior using other database systems, which is why I'm a bit confused.
It's also worth mentioning that my web application's MongoDB connections never fail when connecting to my local development MongoDB server (I assume it would have something to do with the fact that it's a local connection, and that the connection is done through a UNIX socket instead of a network socket, but I could be wrong).
You're misunderstanding AutoReconnect. It is raised when the driver attempts to communicate with the server (to send a command or other operation) and a network failure or similar problem occurs. The name of the exception is meant to communicate that you do not have to create a new instance of MongoClient, the existing client will attempt to reconnect automatically when your application tries the next operation. If the same problem occurs, AutoReconnect is raised again.
I suspect the reason you are seeing sockets timeout (and AutoReconnect being raised) is that there is a load balancer between the server and your application that closes connections after some period of inactivity. For example, this apparently happens on Microsoft's Azure platform after 13 minutes of no activity on a socket. You might be able to fix this by using the socketKeepAlive option, added in PyMongo 2.8. Note that you will also have to set the keepalive interval on your application server to an appropriate value (the default on Linux is 2 hours). See here for more information.

Python Requests Not Cleaning up Connections and Causing Port Overflow?

I'm doing something fairly outside of my comfort zone here, so hopefully I'm just doing something stupid.
I have an Amazon EC2 instance which I'm using to run a specialized database, which is controlled through a webapp inside of Tomcat that provides a REST API. On the same server, I'm running a Python script that uses the Requests library to make hundreds of thousands of simple queries to the database (I don't think it's possible to consolidate the queries, though I am going to try that next.)
The problem: after running the script for a bit, I suddenly get a broken pipe error on my SSH terminal. When I try to log back in with SSH, I keep getting "operation timed out" errors. So I can't even log back in to terminate the Python process and instead have to reboot the EC2 instance (which is a huge pain, especially since I'm using ephemeral storage)
My theory is that each time requests makes a REST call, it activates a pair of ports between Python and Tomcat, but that it never closes the ports when it's done. So python keeps trying to grab more and more ports and eventually either somehow grabs away and locks the SSH port (booting me off), or it just uses all the ports and that causes the network system to crap out somehow (as I said, I'm out of my depth.)
I also tried using httplib2, and was getting a similar problem.
Any ideas? If my port theory is correct, is there a way to force requests to surrender the port when it's done? Or otherwise is there at least a way to tell Ubuntu to keep the SSH port off-limits so that I can at least log back in and terminate the process?
Or is there some sort of best practice to using Python to make lots and lots of very simple REST calls?
Edit:
Solved...do:
s = requests.session()
s.config['keep_alive'] = False
Before making the request to force Requests to release connections when it's done.
My speculation:
https://github.com/kennethreitz/requests/blob/develop/requests/models.py#L539 sets conn to connectionpool.connection_from_url(url)
That leads to https://github.com/kennethreitz/requests/blob/develop/requests/packages/urllib3/connectionpool.py#L562, which leads to https://github.com/kennethreitz/requests/blob/develop/requests/packages/urllib3/connectionpool.py#L167.
This eventually leads to https://github.com/kennethreitz/requests/blob/develop/requests/packages/urllib3/connectionpool.py#L185:
def _new_conn(self):
"""
Return a fresh :class:`httplib.HTTPConnection`.
"""
self.num_connections += 1
log.info("Starting new HTTP connection (%d): %s" %
(self.num_connections, self.host))
return HTTPConnection(host=self.host, port=self.port)
I would suggest hooking a handler up to that logger, and listening for lines that match that one. That would let you see how many connections are being created.
Figured it out...Requests has a default 'Keep Alive' policy on connections which you have to explicitly override by doing
s = requests.session()
s.config['keep_alive'] = False
before you make a request.
From the doc:
"""
Keep-Alive
Excellent news — thanks to urllib3, keep-alive is 100% automatic within a session! Any requests that you make within a session will automatically reuse the appropriate connection!
Note that connections are only released back to the pool for reuse once all body data has been read; be sure to either set prefetch to True or read the content property of the Response object.
If you’d like to disable keep-alive, you can simply set the keep_alive configuration to False:
s = requests.session()
s.config['keep_alive'] = False
"""
There may be a subtle bug in Requests here because I WAS reading the .text and .content properties and it was still not releasing the connections. But explicitly passing 'keep alive' as false fixed the problem.

Categories