I'm using Cloud Run containers to run untrusted (user-supplied) code. The container receives a POST request, runs the code, and responds with the result. For security reasons, it's deployed on a locked down service account, but I also want to block all other network activity. How can this be accomplished?
Cloud Run (managed) currently doesn't offer firewall restrictions to selectively block inbound or outbound traffic by IP/host. I'm assuming you're trying to block connections initiated from container to outside. In the future, Cloud Run has plans to add support for Google Cloud VPC Service Controls feature, so that might help.
However, if you have a chance to use Cloud Run for Anthos (on GKE) which has a similar developer experience but runs on Kubernetes clusters, you can actually easily write Kubernetes NetworkPolicy policies (which I have some recipes here) to control which sort of traffic can come/go from the containers running. You can also use GCE firewall rules and VPC service controls when using a Kubernetes cluster.
Other than that, your only option on a Cloud Run (fully managed) environment is to use Linux iptables command while starting your container to block certain network patterns. Importantly, note that Cloud Run (fully managed) runs on a gVisor sandbox which emulates system calls. And many of the features in iptables are currently not implemented/supported in gVisor. By looking at issue tracker and patches , I can tell that it's on the roadmap and some may even be working today.
You could couple the Cloud Run (managed) deployment to a VPC Network that doesn't have any internet access.
I figured this out for my usecase (blocking all egress).
In the first generation of cloud run atlease, there's 2 eth interfaces - eth0 and eth2. Blocking traffic on eth2 solves egress traffic.
iptables -I OUTPUT -o eth2 -j DROP
Run this on startup of the container/app and then ensure the running application is not run (and hence cannot undo this).
Related
Currently, I have followed the google docs quick start docs for deploying a simple cloud run web server that is connected to AlloyDB. However, in the docs, it all seem to point towards of having to utilize VM for a postgreSQL client, which then is connected to my AlloyDB cluster instance. I believe a connection can only be made within the same VPC and/or a proxy service via the VM(? Please correct me if I'm wrong)
I was wondering, if I only want to give access to services within the same VPC, is having a VM a must? or is there another way?
You're correct. AlloyDB currently only allows connecting via Private IP, so the only way to talk directly to the instances is within the same VPC. The reason all the tutorials (e.g. https://cloud.google.com/alloydb/docs/quickstart/integrate-cloud-run, which is likely the quickstart you mention) talk about a VM is that in order to create your databases themselves within the AlloyDB cluster, set user grants, etc, you need to be able to talk to it from inside the VPC. Another option for example, would be to set up Cloud VPN to some local network to connect your LAN to the VPC directly. But that's slow, costly, and kind of a pain.
Cloud Run itself does not require the VM piece, the quickstart I linked to above walks through setting up the Serverless VPC Connector which is the required piece to connect Cloud Run to AlloyDB. The VM in those instructions is only for configuring the PG database itself. So once you've done all the configuration you need, you can shut down the VM so it's not costing you anything. If you needed to step back in to make configuration changes, you can spin the VM back up, but it's not something that needs to be running for the Cloud Run -> AlloyDB connection.
Providing public ip functionality for AlloyDB is on the roadmap, but I don't have any kind of timeframe for when it will be implemented.
I'd like to have my Google Cloud Run services privately communicate with one another over non-HTTP and/or without having to add bearer authentication in my code.
I'm aware of this documentation from Google which describes how you can do authenticated access between services, although it's obviously only for HTTP.
I think I have a general idea of what's necessary:
Create a custom VPC for my project
Enable the Serverless VPC Connector
What I'm not totally clear on is:
Is any of this necessary? Can Cloud Run services within the same project already see each other?
How do services address one another after this?
Do I gain the ability to use simpler by-convention DNS names? For example, could I have each service in Cloud Run manifest on my VPC as a single first level DNS name like apione and apitwo rather than a larger DNS name that I'd then have to hint in through my deployments?
If not, is there any kind of mechanism for services to discover names?
If I put my managed Cloud SQL postgres database on this network, can I control its DNS name?
Finally, are there any other gotchas I might want to be aware of? You can assume my use case is very simple, two or more long lived services on Cloud Run, doing non-HTTP TCP/UDP communications.
I also found a potentially related Google Cloud Run feature request that is worth upvoting if this isn't currently possible.
Cloud Run services are only reachable through HTTP request. you can't use other network protocol (SSH to log into instances for example, or TCP/UDP communication).
However, Cloud Run can initiate these kind of connection to external services (for instance Compute Engine instances deployed in your VPC, thanks to the serverless VPC Connector).
the serverless VPC connector allow you to make a bridge between the Google Cloud managed environment (where live the Cloud Run (and Cloud Functions/App Engine) instances) and the VPC of your project where you have your own instances (Compute Engine, GKE node pools,...)
Thus you can have a Cloud Run service that reach a Kubernetes pods on GKE through a TCP connection, if it's your requirement.
About service discovery, it's not yet the case but Google work actively on that and Ahmet (Google Cloud Dev Advocate on Cloud Run) has released recently a tool for that. But nothing really build in.
I have a Node API deployed to Google CloudRun and it is responsible for managing external servers (clean, new Amazon EC2 Linux VM's), including through SSH and SFTP. SSH and SFTP actually work eventually but the connections take 2-5 MINUTES to initiate. Sometimes they timeout with handshake timeout errors.
The same service running on my laptop, connecting to the same external servers, has no issues and the connections are as fast as any normal SSH connection.
The deployment on CloudRun is pretty standard. I'm running it with a service account that permits access to secrets, etc. Plenty of memory allocated.
I have a VPC Connector set up, and have routed all traffic through the VPC connector, as per the instructions here: https://cloud.google.com/run/docs/configuring/static-outbound-ip
I also tried setting UseDNS no in the /etc/ssh/sshd_config file on the EC2 as per some suggestions online re: slow SSH logins, but that has not make a difference.
I have rebuilt and redeployed the project a few dozen times and all tests are on brand new EC2 instances.
I am attempting these connections using open source wrappers on the Node ssh2 library, node-ssh and ssh2-sftp-client.
Ideas?
Cloud Run works only until you have a HTTP request active.
You proably don't have an active request during this on Cloud Run, as outside of the active request the CPU is throttled.
Best for this pipeline is Cloud Workflows and regular Compute Engine instances.
You can setup a Workflow to start a Compute Engine for this task, and stop once it finished doing the steps.
I am the author of article: Run shell commands and orchestrate Compute Engine VMs with Cloud Workflows it will guide you how to setup.
Executing the Workflow can be triggered by Cloud Scheduler or by HTTP ping.
I'm new to GCP and trying to make heads and tails of it. So far, I've experienced with GKE and Cloud Run.
In GKE, I can create a Workload (deployment) for a service of any kind under any port I like and allocate resources to it. Then I can create a load balancer and open the ports from the pods to the Internet. The load balancer has an IP that I can use to access the underlying pods.
On the other hand, when I create a Could Run service, I'll give it a docker image and a port and once the service is up and running, it exposes an HTTPS URL! The port that I specify in Cloud Run is the docker's internal port and if I want to access the URL, I have to do that through port 80.
Does this mean that Cloud Run is designed only for HTTP services under port 80? Or maybe I'm missing something?
Technically "no", Cloud Run cannot be used for non-HTTP services. See Cloud Run's container runtime contract.
But also "sort of":
The URL of a Cloud Run service can be kept "private" (and they are by default), this means that nobody but some specific identities are allowed to invoked the Cloud Run service. See this page to learn more)
The container must listen for requests on a certain port, and it does not have CPU outside of request processing. However, it is very easy to wrap your binary into a lightweight HTTP server. See for example the Shell sample that Uses a very small Go HTTP sevrer to invoke an arbitrary shell script.
I recently start using GCP but i have one thing i can't solve.
I have: 1 VM + 1 DB Instance + 1 LB. DB instance allow only conections from the VM IP. bUT THE VM IP allow traffic from all ip (if i configure the firewall to only allow CloudFlare and LB IP's the website crash and refuse conections).
Recently i was under attack, i activate the Cloudflare ddos mode, restart all and in like 6 h the attack come back with the Cloudflare activate. Wen i see mysql conections bump from 20-30 to 254 and all conections are from the IP of the VM so i think the problem are the public accesibility of the VM but i don't know how to solved it...
If i activate my firewall rules to only allow traffic from LB and Cloudflare the web refuses all conections..
Any idea what i can do?
Thanks.
Cloud Support here, unfortunately, we do not have visibility into what is installed on your instance or what software caused the issue.
Generally speaking you're responsible for investigating the source of the vulnerability and taking steps to mitigate it.
I'm writing here some hints that will help you:
Make sure you keep your firewall rules in a sensible manner, e.g. is not a good practice to have a firewall rule to allow all ingress connections on port 22 from all source IPs for obvious reasons.
Since you've already been rooted, change all your passwords: within the Cloud SQL instance, within the GCE instance, even within the GCP project.
It's also a good idea to check who has access to your service accounts, just in case people that aren't currently working for you or your company still have access to them.
If you're using certificates revoke them, generate new ones and share them in a secure way and with the minimum required number of users.
Securing GCE instances is a shared responsability, in general, OWASP hardening guides are really good.
I'm quoting some info here from another StackOverflow thread that might be useful in your case:
General security advice for Google Cloud Platform instances:
Set user permissions at project level.
Connect securely to your instance.
Ensure the project firewall is not open to everyone on the internet.
Use a strong password and store passwords securely.
Ensure that all software is up to date.
Monitor project usage closely via the monitoring API to identify abnormal project usage.
To diagnose trouble with GCE instances, serial port output from the instance can be useful.
You can check the serial port output by clicking on the instance name
and then on "Serial port 1 (console)". Note that this logs are wipped
when instances are shutdown & rebooted, and the log is not visible
when the instance is not started.
Stackdriver monitoring is also helpful to provide an audit trail to
diagnose problems.
You can use the Stackdriver Monitoring Console to set up alerting policies matching given conditions (under which a service is considered unhealthy) that can be set up to trigger email/SMS notifications.
This quickstart for Google Compute Engine instances can be completed in ~10 minutes and shows the convenience of monitoring instances.
Here are some hints you can check on keeping GCP projects secure.