In my project I have several servers which run NodeJS applications using PM2, those were not created by me. I am not that familiar with the PM2. Now I need to start a new server, which is simply a CRON process that queries an ElasticSearch instance.
There are no routes or anything in it, just a CRON with some logging.
Here is my dilemma. I have played with PM2 and I become somewhat familiar with what is it, and what it does. But the question is how shall I run it?
The previous projects do have PM2 config.json with many parameters, and they are started in cluster mode (handled with Nginx), and when I start them I see all process's becoming daemons. But in my case I don't need that. I just need it to run as a single service.
In other words if I use the configuration file to run the PM2, I see it spawned in cluster mode, and it creates chaos as my CRON is fired many times. I don't need that. If I start it in Fork mode, it also spawns instances, but all of them die, except one (due to which they are using same port). I also don't need that.
I just need single service.
I managed to run the my CRON app.js with the singe line, simple as:
PM2 start app.js. It runs in single thread, and I can see it's info with PM2 status. All fine.
If I run it with the single line(as in my case), is it considered ok? Based in my knowledge if I use config.json, it will always run it in fork or cluster.
Is it ok to run it in single line, or do I need still to use a config.json file.
If you only need one process to be run, as is the case, you're doing the right thing.
Related
I am just wondering if I can seperate my pm2 instances. For example, let's say I have one ecosystem file that I start in one terminal, and then I go ahead and start another ecosystem file in another terminal. Both processes seem to merge and start outputting the same logs. Becomes real annoying real quick. Is there a way to have them seperate?
PM2 uses the .pm2 folder by default to place its logs and .pid file. PM2 uses the PM2_HOME environment variable to know where this folder is. I uncovered this when I was moving the folder to a larger drive and ended up with a new instance when I checked on the apps it was monitoring. As stated in the docs about running multiple PM2 instances, you can run multiple PM2 instance simply by specifying a different folder for each:
PM2_HOME='.pm2' pm2 start echo.js --name="echo-node-1"
PM2_HOME='.pm3' pm2 start echo.js --name="echo-node-2"
Keep in mind that to run any PM2 commands, you'll have to prefix with the proper PM2_HOME:
PM2_HOME='.pm2' pm2 list
PM2_HOME='.pm3' pm2 list
I want my testing script to be able to run my server and then kill it at the end of testing. In fact I will probably want to do this mulitple times in the testing script. I tired exec with shelljs. But that seems to me from running any commands after I start my server for testing it also seems like it would be hard to find the sever to kill it after starting it (especially on different OSes). How might I do such a thing?
On different OS you will need to find processId to kill it. This is tedious and not recommended process.
Use a wrapper around like pm2 or nodemon for servers.
In testing case you can make use of process.exit(0) and expose it in your make code.
I've searched a lot to figure out this question, but I didn't get clear explanation. Is there only one difference thing that clustered app can be scaled out and forked app cannot be?
PM2's public site explains Cluster mode can do these feature but no one says about pros of Fork mode (maybe, it can get NODE_APP_INSTANCE variable).
I feel like Cluster might be part of Fork because Fork seems like to be used in general. So, I guess Fork means just 'forked process' from the point of PM2 and Cluster means 'forked process that is able to be scaled out'. Then, why should I use Fork mode?
The main difference between fork_mode and cluster_mode is that it orders pm2 to use either the child_process.fork api or the cluster api.
What does this means internally?
Fork mode
Take the fork mode as a basic process spawning. This allows to change the exec_interpreter, so that you can run a php or a python server with pm2. Yes, the exec_interpreter is the "command" used to start the child process. By default, pm2 will use node so that pm2 start server.js will do something like:
require('child_process').spawn('node', ['server.js'])
This mode is very useful because it enables a lot of possibilities. For example, you could launch multiple servers on pre-established ports which will then be load-balanced by HAProxy or Nginx.
Cluster mode
The cluster will only work with node as it's exec_interpreter because it will access to the nodejs cluster module (eg: isMaster, fork methods etc.). This is great for zero-configuration process management because the process will automatically be forked in multiple instances.
For example pm2 start -i 4 server.js will launch 4 instances of server.js and let the cluster module handle load balancing.
Node.js is single-thread.
That means only 1 core of your Intel quad-core CPU can execute the node application.
It called: fork_mode.
We use it for local dev.
pm2 start server.js -i 0 helps you running 1 node thread on each core of your CPU.
And auto-load-balance the stateless coming requests.
On the same port.
We call it: cluster_mode.
Which is used for the sake of performance on production.
You may also choose to do this on local dev if you want to stress test your PC :)
Documentation and sources are really misleading here.
Reading up on this in the sources, the only differences seems to be, that they use either node cluster or child_process API. Since cluster uses the latter, you are actually doing the same. There is just a lot more custom stdio passing around happening inn fork_mode. Also cluster can only be communicated with via strings, not objects.
By default you are using fork_mode. If you pass the the -i [number]-option, you're going into cluster_mode, which you generally aim for w/ pm2.
Also fork_mode instance probably can't listen on the same port due to EADDRINUSE. cluster_mode can. This way you also can structure you app to run on the same port being automatically load balanced. You have to build apps without state then though e.g. sessions, dbs.
What is the easiest way to achieve zero downtime for my Node.js application?
I have an app that requires the following steps for redeployment:
npm install
node_modules/.bin/bower install
node_modules/.bin/gulp
The result of these operations is the ready-to-run application in the generated by the gulpfile.js directory named build. In this directory I have a currently-running instance of the same application (currently launched via forever like this -- forever start server.js).
As far as I know, it is not possible to achieve zero downtime via forever module, so I decided to choose another way to do it.
I saw pm2 but I found it very complex tbh (prove me wrong if you don't feel the same).
I also saw naught but I can't even start my application via naught start server.js -- it doesn't even print anything in stdout / stderr.
I also saw up-time but I didn't get the idea -- how will it handle situation when I run gulp that should replace files in the directory where currently-running instance work at the moment?
Regarding of handling replaced files during build: if these files is used by Node.js app then all changes will be applied upon process restart (since these files are loaded into memory), browser frontend files could also be cached in application memory to achieve similar behavior (changes applied only upon restart or/and cache invalidation).
We're using pm2 in cluster mode.
pm2 start app.js -i
The above command starts app.js in cluster mode on all available CPU cores.
zero downtime restart:
pm2 gracefulReload all
this command restarts all processes sequentially, so if you have more than one process up and running there is always at least one process that servers requests during restart.
If you have only one process of app.js you can start it in cluster mode and run pm2 scale app.js 2 (starts one more process) then pm2 gracefulReload all and then pm2 scale app.js 1 (removes previously started process).
Though I think app restarting is not main problem of zero downtime deployment, we've not managed to handle DB migrations, so full app shutdown is needed to apply DB changes. Also there could be an issue with browser frontend files when during deploy user obtained the new version of them, but AJAX request is processed by old version of server process, in this case sticky sessions and API versioning came to the rescue.
My application is built with three distincts servers: each one of them serves a different purpose and they must stay separated (at least, for using more than one core). As an example (this is not the real thing) you could think about this set up as one server managing user authentication, another one serving as the game engine, another one as a pubsub server. Logically the "application" is only one and clients connect to one or another server depending on their specific need.
Now I'm trying to figure out the best way to run a setup like this in a production environment.
The simplest way could be to have a bash script that would run each server in background one after the other. One problem with this approach would be that in the case I need to restart the "application", I should have saved each server's pid and kill each one.
Another way would be to use a node process that would run each servers as its own child (using child_process.spawn). Node spawning nodes. Is that stupid for some reason? This way I'd have a single process to kill when I need to stop/restart the whole application.
What do you think?
If you're on Linux or another *nix OS, you might try writing an init script that start/stop/restart your application. here's an example.
Use specific tools for process monitoring. Monit for example can monitor your processes by their pid and restart them whenever they die, and you can manually restart each process with the monit-cmd or with their web-gui.
So in your example you would create 3 independent processes and tell monit to monitor each of them.
I ended up creating a wrapper/supervisor script in Node that uses child_process.spawn to execute all three processes.
it pipes each process stdout/stderr to the its stdout/stderr
it intercepts errors of each process, logs them, then exit (as it were its fault)
It then forks and daemonize itself
I can stop the whole thing using the start/stop paradigm.
Now that I have a robust daemon, I can create a unix script to start/stop it on boot/shutdown as usual (as #Levi says)
See also my other (related) Q: NodeJS: will this code run multi-core or not?