ConCoord is a novel coordination service that provides replication and synchronization support for large-scale distributed systems. ConCoord employs an object-oriented approach, in which the system creates and maintains live replicas for Python objects written by the user. ConCoord converts these Python objects into Paxos Replicated State Machines (RSM) and enables clients to do method invocations on them transparently as if they are local objects. ConCoord uses these replicated objects to implement coordination and synchronization constructs in large-scale distributed systems, in effect establishing a transparent way of providing a coordination service.
| Authors: |
|
|---|---|
| Version: | 1.1.0 |
| Date: | 2014-04-03 |
The minimum requirements for ConCoord are:
- python 2.7.2 or later - dnspython-1.9.4 - msgpack-python
ConCoord can be installed from source with:
$ python setup.py install
ConCoord is also available for install through PyPI:
$ pip install concoord
To use ConCoord you need a Python object that can be used for the coordination of your distributed system. In the ConCoord distribution, we offer ready-to-use objects that cover the most common coordination needs. So first, let's start a ConCoord instance with an object in the distribution, namely Counter under concoord/object/counter.py.
To distribute the local object you should start at least one replica.
To start a bootstrap replica node that doesn't need to be connected to another replica, use the following command:
$ concoord replica -o concoord.object.counter.Counter -a 127.0.0.1 -p 14000
To start replica nodes to join an active ConCoord instance, use the following command to connect to another replica:
$ concoord replica -o concoord.object.counter.Counter -b 127.0.0.1:14000 -a 127.0.0.1 -p 14001
You can specify the port and the address of any replica with options -p and -a respectively. The replicas can also be run in the debug mode or with a logger with the commands shown below:
Usage: concoord replica [-h] [-a ADDR] [-p PORT] [-b BOOTSTRAP]
[-o OBJECTNAME] [-l LOGGER] [-n DOMAIN]
[-r] [-w] [-d]
- where,
-h, --help show this help message and exit -a ADDR, --addr ADDR address for the node -p PORT, --port PORT port for the node -b BOOTSTRAP, --boot BOOTSTRAP address:port tuple for the bootstrap peer -o OBJECTNAME, --objectname OBJECTNAME client object dotted name -l LOGGER, --logger LOGGER logger address -n DOMAIN, --domainname DOMAIN domain name that the name server will accept queries for -r, --route53 use Route53 -w, --writetodisk writing to disk on/off -d, --debug debug on/off
You can dynamically locate nodes in a given ConCoord instance using DNS queries if the instance includes replicas that can act as name servers. There are two ways you can run a ConCoord Replica as a name server.
- Master Name Server: Keeps track of the view and responds to DNS queries itself. Requires su privileges to bind to Port 53.
- Route53 Name Server: Keeps track of the view and updates an Amazon Route53 account. Amazon Route53 answers to DNS queries on behalf of the slave name server. Requires a ready-to-use Amazon Route53 account.
To use a replica node as a master name server first you have to setup the name server delegations (you can do this by updating the domain name server information of any domain name you own from the domain registrar you use (godaddy, namecheap etc.)). Once all the delegations are setup for the ip address the replica uses, you can start a replica node as a name server for counterdomain.com as follows:
$ sudo concoord replica -o concoord.object.counter.Counter -a 127.0.0.1 -n counterdomain.com
And to start the replica to join an already running ConCoord instance, provide the bootstrap:
$ sudo concoord replica -o concoord.object.counter.Counter -a 127.0.0.1 -b 127.0.0.1:14000 -n counterdomain.com
When the replica starts running, you can send queries for counterdomain.com and see the most current set of nodes as follows:
$ dig -t a counterdomain.com # returns set of Replicas $ dig -t srv _concoord._tcp.counterdomain.com # returns set of Replicas with ports $ dig -t txt counterdomain.com # returns set of all nodes $ dig -t ns counterdomain.com # returns set of name servers
If you want to run the name server without proper delegation setup, you can query the name server bound to 127.0.0.1 specifically as follows:
$ dig -t txt counterdomain.com @127.0.0.1 # returns set of all nodes
Note that this would only work for a, srv and txt queries, since ns queries require absolute dns names or origins, not an ip address.
First make sure that boto is installed on the machine you want to run the Route53 name server. You can easily install boto as follows:
$ pip install boto
Before starting a name server connected to Amazon Route 53, you should have a Route53 account set up and ready to receive requests. This is done through the AWS Console (http://console.aws.amazon.com/route53), by creating a new Hosted Zone to host your domain name.
After your Route53 account is set up, the name server can update Route53 records every time the view of the system changes.
To use the Name Server to update Amazon Route53, you should provide your AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY. You can retrieve these from the AWS Console (http://console.aws.amazon.com/iam), by looking under the security credentials of the username that you used while creating the Hosted Zone for your domain name. Once you have the information, you can set up ConCoord configuration easily as follows:
$ concoord route53id [AWS_ACCESS_KEY_ID] $ concoord route53key [AWS_SECRET_ACCESS_KEY]
Once you make sure that your Route53 account is set up and the configuration file includes your AWS credentials, you can start the replica with a name server as follows:
$ concoord replica -o concoord.object.counter.Counter -n counterdomain.com -r
When the replica starts running, you can send queries for counterdomain.com and see the most current set of nodes as follows:
$ dig -t a counterdomain.com # returns set of Replicas $ dig -t srv _concoord._tcp.counterdomain.com # returns set of Replicas with ports $ dig -t txt counterdomain.com # returns set of all nodes $ dig -t ns counterdomain.com # returns set of name servers
Once you have a ConCoord instance running with your object, it is easy to access your object.
The proxy for the Counter object is also included in the distribution. You can import and use this proxy object in your code. Depending on how you set your name server up, you can access your object with the ipaddr:port pair or the domainname. In the example below, the ipaddr:port of both replica nodes are used. As a result, the client will be able to do method invocations on the object as long as one of the replicas is alive:
>>> from concoord.proxy.counter import Counter
>>> c = Counter("127.0.0.1:14000, 127.0.0.1:14001")
>>> c.increment()
>>> c.increment()
>>> c.getvalue()
2
At any point to reinitialize an object after it is deployed on replicas, you should call __concoordinit__ function:
>>> from concoord.proxy.counter import Counter
>>> c = Counter("127.0.0.1:14000, 127.0.0.1:14001")
>>> c.increment()
>>> c.__concoordinit__()
>>> c.increment()
>>> c.getvalue()
1
OpenReplica provides easily launch for concoord on remote machines, and it is especially built to launch concoord instances on Amazon EC2 servers easily. In this section we cover how you can setup a set of machines and launch concoord remotely using OpenReplica.
You can easily launch your EC2 instances using the web interface provided by Amazon AWS. In this example we launch an instance running 64-bit Amazon Linux. Before you launch your instance, you will have to create a keypair to login to your instances, make sure to keep your private key safe. The best way is to move your key-pair to the .ssh directory once you download the [my-key-pair].pem file:
$ mv [my-key-pair].pem ~/.ssh/ $ ssh-add ~/.ssh/[my-key-pair].pem
Once you did these, now is the time to connect to your instance:
$ ssh -i [my-key-pair].pem ec2-user@[public_dns_name]
After this point on, you should be able to connect to your instance without explicitly passing the key as a parameter, as follows:
$ ssh ec2-user@[public_dns_name]
To enable execution of remote sudo commands SSH into your instance and run:
$ sudo visudo
In the file that opens find the line 'Defaults requiretty' and add line 'Defaults:ec2-user !requiretty' below it. Save the file and exit. The file opens with the vi editor. If you are not familiar with vi commands, to insert new text press i to go into insert mode and add the new line. After you are done, press ESC and ZZ to save and exit.
At this point, your EC2 instance should be easily accessible.
Using OpenReplica developers can configure and remotely launch ConCoord instances on remote machines easily. Firstly, OpenReplica keeps the configuration information of machines that will be running the ConCoord instance. Secondly, using ConCoord developers can easily start replica and nameserver nodes on the machines that are added to the configuration. And lastly, using OpenReplica developers can easily manage EC2 instances.
Like ConCoord, OpenReplica can be used as a commandline script. It supports the following commands:
openreplica config - prints config file openreplica addsshkey [sshkeyfilename] - adds sshkey filename to config openreplica addusername [ssh username] - adds ssh username to config openreplica addnode [publicdns] - adds public dns for node to config openreplica setup [publicdns] - downloads and sets up concoord on the given node openreplica replica [concoord arguments] - starts a replica node on EC2 openreplica nameserver [concoord arguments] - starts a nameserver node on EC2
To use OpenReplica, you should first register the nodes you want to use. To register nodes, you will need the filename of the sshkey you use to ssh into these nodes, as well as the username.
To add the sshkey:
$ openreplica addsshkey [my-key-pair].pem
To add the username:
$ openreplica addusername ec2-user
To add a node:
$ openreplica addnode [public_dns_name]
When adding nodes to OpenReplica, it automatically checks the nodes for eligibility to run ConCoord and warns the user if an update or change is required. Similarly, if ConCoord cannot connect to the node, it lets the user know:
$ openreplica addsshkey concoord.pem Adding SSHKEY to CONFIG: concoord.pem $ openreplica addusername ec2-user Adding USERNAME to CONFIG: ec2-user $ openreplica addnode ec2-54-186-26-155.us-west-2.compute.amazonaws.com Adding NODE to CONFIG: ec2-54-186-26-155.us-west-2.compute.amazonaws.com Cannot connect to node, check if it is up and running.
Once OpenReplica is set up, nodes can be started as if they are being started on the local machine.
To start a bootstrap replica node that doesn't need to be connected to another replica:
$ openreplica replica -o concoord.object.counter.Counter -a 127.0.0.1 -p 14000
To start replica nodes to join an active ConCoord instance:
$ openreplica replica -o concoord.object.counter.Counter -b 127.0.0.1:14000 -a 127.0.0.1 -p 14001
The nodes can also be run in the debug mode or with a logger with the commands shown below:
Usage: openreplica replica [-h] [-a ADDR] [-p PORT] [-b BOOTSTRAP]
[-o OBJECTNAME] [-l LOGGER] [-n DOMAIN]
[-r] [-w] [-d]
- where,
-h, --help show this help message and exit -a ADDR, --addr ADDR address for the node -p PORT, --port PORT port for the node -b BOOTSTRAP, --boot BOOTSTRAP address:port tuple for the bootstrap peer -o OBJECTNAME, --objectname OBJECTNAME client object dotted name -l LOGGER, --logger LOGGER logger address -n DOMAIN, --domainname DOMAIN domain name that the name server will accept queries for -r, --route53 use Route53 -w, --writetodisk writing to disk on/off -d, --debug debug on/off
You can dynamically locate nodes in a given ConCoord instance using DNS queries if the instance includes replicas that can act as name servers. There are two ways you can run a ConCoord Replica as a name server.
- Master Name Server: Keeps track of the view and responds to DNS queries itself. Requires su privileges to bind to port 53.
- Route53 Name Server: Keeps track of the view and updates an Amazon Route53 account. Amazon Route53 answers to DNS queries on behalf of the slave name server. Requires a ready-to-use Amazon Route53 account.
To use a replica node as a master name server first you have to setup the name server delegations (you can do this by updating the domain name server information of any domain name you own from the domain registrar you use (godaddy, namecheap etc.)). Once all the delegations are setup for the ip address the replica uses, you can start a replica node as a name server for counterdomain.com as follows:
$ openreplica replica -o concoord.object.counter.Counter -a 127.0.0.1 -n counterdomain.com
And to start the replica to join an already running ConCoord instance, provide the bootstrap:
$ openreplica replica -o concoord.object.counter.Counter -a 127.0.0.1 -b 127.0.0.1:14000 -n counterdomain.com
When the replica starts running, you can send queries for counterdomain.com and see the most current set of nodes as follows:
$ dig -t a counterdomain.com # returns set of Replicas $ dig -t srv _concoord._tcp.counterdomain.com # returns set of Replicas with ports $ dig -t txt counterdomain.com # returns set of all nodes $ dig -t ns counterdomain.com # returns set of name servers
First make sure that boto is installed on the machine you want to run the Route53 name server. OpenReplica tries to do this automatically when a replica is run as a Route53 name server, but if it fails to do so, you can easily install boto on the machine you want as follows:
$ pip install boto
Before starting a name server connected to Amazon Route53, you should have a Route53 account set up and ready to receive requests. This is done through the AWS Console (http://console.aws.amazon.com/route53), by creating a new Hosted Zone to host your domain name.
After your Route53 account is set up, the name server can update Route53 records every time the view of the system changes.
To use the Name Server to update Amazon Route53, you should provide your AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY. You can retrieve these from the AWS Console (http://console.aws.amazon.com/iam/), by looking under the security credentials of the username that you used while creating the Hosted Zone for your domain name. Once you have the information, you can set up Route53 configuration easily as follows:
$ openreplica route53 [public_dns AWS_ACCESS_KEY_ID AWS_SECRET_ACCESS_KEY]
Once you make sure that your Route53 account is set up and the configuration file includes your AWS credentials, you can start the replica with a name server as follows:
$ openreplica replica -o concoord.object.counter.Counter -n counterdomain.com -r
When the replica starts running, you can send queries for counterdomain.com and see the most current set of nodes as follows:
$ dig -t a counterdomain.com # returns set of Replicas $ dig -t srv _concoord._tcp.counterdomain.com # returns set of Replicas with ports $ dig -t txt counterdomain.com # returns set of all nodes $ dig -t ns counterdomain.com # returns set of name servers
For cases when the objects included in the ConCoord distribution do not meet your coordination needs, ConCoord lets you convert your local Python objects into distributable objects very easily.
To walk through the ConCoord approach, you will use a different Counter object that increments and decrements by ten, namely tencounter.py. Once you install ConCoord, you can create coordination objects and save them anywhere in your filesystem. After you create tencounter.py, you can save tencounter.py under /foo/tencounter.py:
class TenCounter:
def __init__(self, value=0):
self.value = value
def decrement(self):
self.value -= 10
def increment(self):
self.value += 10
def getvalue(self):
return self.value
def __str__(self):
return "The counter value is %d" % self.value
Once you have created the object, update your PYTHONPATH accordingly, so that the objects can be found and imported:
$ export PYTHONPATH=$PYTHONPATH:/foo/
Clients will use a proxy object to do method calls on the object. To create the proxy object automatically, you can use the following command:
$ concoord object -o tencounter.TenCounter
Usage: concoord object [-h] [-o OBJECTNAME] [-t SECURITYTOKEN] [-p PROXYTYPE]
[-s] [-v]
- where,
-h, --help show this help message and exit -o OBJECTNAME, --objectname OBJECTNAME client object dotted name module.Class -t SECURITYTOKEN, --token SECURITYTOKEN security token -p PROXYTYPE, --proxytype PROXYTYPE 0:BASIC, 1:BLOCKING, 2:CLIENT-SIDE BATCHING, 3:SERVER-SIDE BATCHING -s, --safe safety checking on/off -v, --verbose verbose mode on/off
This script will create a proxy file under the directory that the object resides (i.e. /foo/):
/foo/tencounterproxy.py := the proxy that can be used by the client
IMPORTANT NOTE: ConCoord objects treat __init__ functions specially in two ways:
1) When Replicas go live, the object is instantiated calling the __init__ without any arguments. Therefore, while implementing coordination objects, the __init__ method should be implemented to be able to run without explicit arguments. You can use defaults to implement an __init__ method that accepts arguments.
2) In the proxy created, the __init__ function is used to initialize the Client-Replica connection. This way, multiple clients can connect to a ConCoord instance without reinitializing the object. During proxy generation, the original __init__ function is renamed as __concoordinit__, to reinitialize the object the user can call __concoordinit__ at any point.
After this point on, you can use TenCounter just like we use Counter before.
You can create bundles to use at the server and client sides using the Makefile provided. Remember to add the objects you have created in these bundles.
To create a bundle that can run replicas:
$ make server
To create a bundle that can run a client and connect to an existing ConCoord instance:
$ make client
We have two kinds of loggers for ConCoord:: * Console Logger * Network Logger
Both of these loggers are included under utils.py. To start the NetworkLogger, use the logdaemon.py on the host you want to keep the Logger.
ConCoord provides a distributed and fault-tolerant threading library. The library includes:
- Lock
- RLock
- Semaphore
- BoundedSemaphore
- Barrier
- Condition
The implementations of distributed synchronization objects follow the implementations in the Python threading library. We will walk through an example below using the Semaphore object under concoord/object/semaphore.py
In the blocking object implementation, the method invocations that use an object from the threading library requires an extra argument _concoord_command. This argument is used by the calling Replica node to relate any blocking/unblocking method invocation to a specific client. This way, even if the client disconnects and reconnects, the ConCoord instance will remain in a safe state:
from concoord.threadingobject.dsemaphore import DSemaphore
class Semaphore:
def __init__(self, count=1):
self.semaphore = DSemaphore(count)
def __repr__(self):
return repr(self.semaphore)
def acquire(self, _concoord_command):
try:
return self.semaphore.acquire(_concoord_command)
except Exception as e:
raise e
def release(self, _concoord_command):
try:
return self.semaphore.release(_concoord_command)
except Exception as e:
raise e
def __str__(self):
return str(self.semaphore)
To create the proxy for this blocking object we will use the following command:
$ concoord object -o concoord.object.semaphore.Semaphore -p 1
This command creates the proxy that supports blocking operations. Now you can use blocking objects just like basic ConCoord objects. First, we start the replica nodes the same way we did before as follows:
$ concoord replica -o concoord.object.semaphore.Semaphore -a 127.0.0.1 -p 14000
To test the functionality, you can use multiple clients or print out the Semaphore object as follows:
>>> from semaphoreproxy import Semaphore
>>> s = Semaphore("127.0.0.1:14000")
>>> s.acquire()
True
>>> i = 10
>>> i += 5
>>> s
<DSemaphore count=0>
>>> s.release()
>>> s
<DSemaphore count=1>
>>>
Visit http://openreplica.org to get more information on ConCoord.
If you believe you have found a bug or have a problem you need assistance with, you can get in touch with us by emailing concoord@systems.cs.cornell.edu