[RFC] Github #329: Kimchi must not run as root
Information about the issue: https://github.com/kimchi-project/kimchi/issues/329 We had some ideas about it in the last weekly scrum: - creating an user 'kimchi' with a lot of privileges. It is the simplest of the solutions but implies in a lot of annoyances (different system paths between distros, libvirt does not work the same way in all distros). There is no tell about the amount of new bugs and issues that shall emerge from such change. - Separating the UI and the backend. This is my personal favourite but I believe it is also the hardest. We can implement this by separating frontend and backend as 2 separate cherrypy processes, the backend runs as root and the frontend runs as a regular user. The communication would be done using the REST API. The other approach would be the backend running as a regular python daemon, with root privileges, and kimchi would communicate with it using RPC. I believe the latter is more elegant and the former is easier to implement. Things to consider: - Distro support: Ubuntu, for example, behaves very different from Fedora as far as libvirt is concerned. The packaging model (apt-get instead of yum) differs in support as well. I think it's a fair guess that RHEL 6 and Suse will have different behavior as well. - VM visibility: in the first idea (different user) only the VMs created by this specific user would be visible to kimchi, unless we do something about it (tweaking libvirt configuration perhaps?). - User authentication. Right now the user authentication presented in kimchi exists simply to authenticate it as a regular user of the host. The owner of the process will be root, doesn't matter which user logs in (and I guess this is the critical security flaw we have). Do we need ro rethink the authentication model as well? Please provide your input and ideas! Daniel
On 03/05/2014 11:49 AM, Daniel H Barboza wrote:
Information about the issue:
https://github.com/kimchi-project/kimchi/issues/329
We had some ideas about it in the last weekly scrum:
- creating an user 'kimchi' with a lot of privileges. It is the simplest of the solutions but implies in a lot of annoyances (different system paths between distros, libvirt does not work the same way in all distros). There is no tell about the amount of new bugs and issues that shall emerge from such change.
It will be hard to implement as we use some python bindings - libvirt, yum, apt... For libvirt we could use libvrt.openAuth() using a sudo user But how about yum/apt which requires root privileges to run?
- Separating the UI and the backend. This is my personal favourite but I believe it is also the hardest. We can implement this by separating frontend and backend as 2 separate cherrypy processes, the backend runs as root and the frontend runs as a regular user. The communication would be done using the REST API. The other approach would be the backend running as a regular python daemon, with root privileges, and kimchi would communicate with it using RPC. I believe the latter is more elegant and the former is easier to implement.
I personally tend to this solution as well. My first attempt would be try to run everything under /model and mockmodel (which is the real backend code) as root and start the cherrypy server as normal user. We need to verify if cherrypy provides some built-in solution for it what can help us to solve this problem.
Things to consider:
- Distro support: Ubuntu, for example, behaves very different from Fedora as far as libvirt is concerned. The packaging model (apt-get instead of yum) differs in support as well. I think it's a fair guess that RHEL 6 and Suse will have different behavior as well.
- VM visibility: in the first idea (different user) only the VMs created by this specific user would be visible to kimchi, unless we do something about it (tweaking libvirt configuration perhaps?).
It is not true. All the kimchi requests will be performed in the same libvirt connection created for the "kimchi" user. So instead of manage all the vms on root user it will be on kimchi user.
- User authentication. Right now the user authentication presented in kimchi exists simply to authenticate it as a regular user of the host. The owner of the process will be root, doesn't matter which user logs in (and I guess this is the critical security flaw we have). Do we need ro rethink the authentication model as well?
I can't find a connection between those things. Independent who is running the process (root or kimchi user) the authentication will be based on system configuration (PAM auth)
Please provide your input and ideas!
Daniel
_______________________________________________ Kimchi-devel mailing list Kimchi-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/kimchi-devel
Talk with Mark about the separation of frontend/backend. Thanks for him to contribute the following idea: We think even for backend, not all codes need to run as root, such as iso search, directory creation and so on. Gurentee safe may be minimize of use of root. For previledge cmd: run as super user and limit cmds needed to a single module. For libvirt connection: Enable libvirt connection authentication, we can add sasl user in libvirt for kimchi. On 2014年03月05日 22:49, Daniel H Barboza wrote:
Information about the issue:
https://github.com/kimchi-project/kimchi/issues/329
We had some ideas about it in the last weekly scrum:
- creating an user 'kimchi' with a lot of privileges. It is the simplest of the solutions but implies in a lot of annoyances (different system paths between distros, libvirt does not work the same way in all distros). There is no tell about the amount of new bugs and issues that shall emerge from such change.
- Separating the UI and the backend. This is my personal favourite but I believe it is also the hardest. We can implement this by separating frontend and backend as 2 separate cherrypy processes, the backend runs as root and the frontend runs as a regular user. The communication would be done using the REST API. The other approach would be the backend running as a regular python daemon, with root privileges, and kimchi would communicate with it using RPC. I believe the latter is more elegant and the former is easier to implement.
Things to consider:
- Distro support: Ubuntu, for example, behaves very different from Fedora as far as libvirt is concerned. The packaging model (apt-get instead of yum) differs in support as well. I think it's a fair guess that RHEL 6 and Suse will have different behavior as well.
- VM visibility: in the first idea (different user) only the VMs created by this specific user would be visible to kimchi, unless we do something about it (tweaking libvirt configuration perhaps?).
- User authentication. Right now the user authentication presented in kimchi exists simply to authenticate it as a regular user of the host. The owner of the process will be root, doesn't matter which user logs in (and I guess this is the critical security flaw we have). Do we need ro rethink the authentication model as well?
Please provide your input and ideas!
Daniel
_______________________________________________ Kimchi-devel mailing list Kimchi-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/kimchi-devel
On 03/05/2014 10:49 PM, Daniel H Barboza wrote:
Information about the issue:
https://github.com/kimchi-project/kimchi/issues/329
We had some ideas about it in the last weekly scrum:
- creating an user 'kimchi' with a lot of privileges. It is the simplest of the solutions but implies in a lot of annoyances (different system paths between distros, libvirt does not work the same way in all distros). There is no tell about the amount of new bugs and issues that shall emerge from such change.
- Separating the UI and the backend. This is my personal favourite but I believe it is also the hardest. We can implement this by separating frontend and backend as 2 separate cherrypy processes, the backend runs as root and the frontend runs as a regular user. The communication would be done using the REST API. The other approach would be the backend running as a regular python daemon, with root privileges, and kimchi would communicate with it using RPC. I believe the latter is more elegant and the former is easier to implement.
from the cherrypy doc, we can customized the components of cherrypy flexibly. Not sure, we can customized WSGI components or other components. And the let cherrypy daemon start with root, and it fork one process with root user, let this process to handle the libvirt API or other API need root privilege. The father cherrypy daemon then changes him self user to "kimchi". It handle other request does not need root privilege. Maybe we need process communication between these processes. That's need to look into the CherryPy mechanism to see the feasibility. [DOC]: Finally, there is the CORE LAYER, which uses the core API's to construct the default components which are available at higher layers. You can think of the default components as the 'reference implementation' for CherryPy. Megaframeworks (and advanced users) may replace the default components with customized or extended components. The core API's are: * Application API * Engine API * Request API * Server API * WSGI API These API's are described in the CherryPy specification:
Things to consider:
- Distro support: Ubuntu, for example, behaves very different from Fedora as far as libvirt is concerned. The packaging model (apt-get instead of yum) differs in support as well. I think it's a fair guess that RHEL 6 and Suse will have different behavior as well.
- VM visibility: in the first idea (different user) only the VMs created by this specific user would be visible to kimchi, unless we do something about it (tweaking libvirt configuration perhaps?).
- User authentication. Right now the user authentication presented in kimchi exists simply to authenticate it as a regular user of the host. The owner of the process will be root, doesn't matter which user logs in (and I guess this is the critical security flaw we have). Do we need ro rethink the authentication model as well?
Please provide your input and ideas!
Daniel
_______________________________________________ Kimchi-devel mailing list Kimchi-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/kimchi-devel
-- Thanks and best regards! Sheldon Feng(冯少合)<shaohef@linux.vnet.ibm.com> IBM Linux Technology Center
I've tested libvirt Connection authentication in RHEL6.5, Ubuntu and Fedora 19. This is the link I've used as guide: http://libvirt.org/auth.html#ACL_server_unix_perms The idea was to make a simple sanity check to verify libvirt connectivity using the python binding. In kimchi's root dir: $ PYTHONPATH=src python Python 2.7.5 (default, Nov 12 2013, 16:18:42) [GCC 4.8.2 20131017 (Red Hat 4.8.2-1)] on linux2 Type "help", "copyright", "credits" or "license" for more information.
from kimchi.model.libvirtconnection import LibvirtConnection libvirt_conn = LibvirtConnection('qemu:///system') conn = libvirt_conn.get()
from kimchi.model.libvirtconnection import LibvirtConnection libvirt_conn = LibvirtConnection('qemu:///system') conn = libvirt_conn.get()
This test was successful in Fedora 19 (Lenovo T410 laptop) and in Ubuntu 12.04.3 LTS running in a Intel server. This test FAILED in RHEL6.5 running in Intel. In the RHEL6.5 test I've tried leaving the socket open to R+W to all (777), which in theory would let anyone open a libvirt connection regardless of group permission. Still no good. A non-root user couldn't open a libvirt connection: $ PYTHONPATH=src python Python 2.6.6 (r266:84292, Nov 21 2013, 10:50:32) [GCC 4.4.7 20120313 (Red Hat 4.4.7-4)] on linux2 Type "help", "copyright", "credits" or "license" for more information. libvir: XML-RPC error : authentication failed: Authorization requires authentication but no agent is available. Unable to connect to libvirt. libvir: XML-RPC error : authentication failed: Authorization requires authentication but no agent is available. Unable to connect to libvirt. This test runs with no error if using "sudo" to launch the python interpreter. I haven't tested in Suse, but I believe these tests I've made are enough to show that we might run into issues if we rely solely in the functioning of libvirt across all the distros kimchi supports. A solution for that is running all related libvirt calls with root. I agree with Royce and Mark that we don't need to run the entire backend as root, but for this first interaction I think it's simpler to run all the backend as root. About the frontend/backend separation, I'll look into what Sheldon said. Perhaps cherrypy can provide an easier way to communicate both parts. Daniel On 03/06/2014 04:51 AM, Sheldon wrote:
On 03/05/2014 10:49 PM, Daniel H Barboza wrote:
Information about the issue:
https://github.com/kimchi-project/kimchi/issues/329
We had some ideas about it in the last weekly scrum:
- creating an user 'kimchi' with a lot of privileges. It is the simplest of the solutions but implies in a lot of annoyances (different system paths between distros, libvirt does not work the same way in all distros). There is no tell about the amount of new bugs and issues that shall emerge from such change.
- Separating the UI and the backend. This is my personal favourite but I believe it is also the hardest. We can implement this by separating frontend and backend as 2 separate cherrypy processes, the backend runs as root and the frontend runs as a regular user. The communication would be done using the REST API. The other approach would be the backend running as a regular python daemon, with root privileges, and kimchi would communicate with it using RPC. I believe the latter is more elegant and the former is easier to implement.
from the cherrypy doc, we can customized the components of cherrypy flexibly.
Not sure, we can customized WSGI components or other components.
And the let cherrypy daemon start with root, and it fork one process with root user, let this process to handle the libvirt API or other API need root privilege.
The father cherrypy daemon then changes him self user to "kimchi". It handle other request does not need root privilege.
Maybe we need process communication between these processes.
That's need to look into the CherryPy mechanism to see the feasibility.
[DOC]: Finally, there is the CORE LAYER, which uses the core API's to construct the default components which are available at higher layers. You can think of the default components as the 'reference implementation' for CherryPy. Megaframeworks (and advanced users) may replace the default components with customized or extended components. The core API's are:
* Application API * Engine API * Request API * Server API * WSGI API
These API's are described in the CherryPy specification:
Things to consider:
- Distro support: Ubuntu, for example, behaves very different from Fedora as far as libvirt is concerned. The packaging model (apt-get instead of yum) differs in support as well. I think it's a fair guess that RHEL 6 and Suse will have different behavior as well.
- VM visibility: in the first idea (different user) only the VMs created by this specific user would be visible to kimchi, unless we do something about it (tweaking libvirt configuration perhaps?).
- User authentication. Right now the user authentication presented in kimchi exists simply to authenticate it as a regular user of the host. The owner of the process will be root, doesn't matter which user logs in (and I guess this is the critical security flaw we have). Do we need ro rethink the authentication model as well?
Please provide your input and ideas!
Daniel
_______________________________________________ Kimchi-devel mailing list Kimchi-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/kimchi-devel
(ps: this is the same comment I've made in github https://github.com/kimchi-project/kimchi/issues/329) I've been working in this issue for about 10 days. The approach I've chosen was to use RPC to separate the backend (the 'model' and 'mockmodel' classes) running as root from the UI running as regular user. If done right, the frontend code will use the model classes like it always did, but these classes will have root privilege because they were instanced by the RPC server (which runs as root). Now I'll explain why it didn't work so far, lol. The RPC I've been using is called PyRO (https://pypi.python.org/pypi/Pyro4). It's rather simple to make it work and at first solved the problem. As far as I could tell the UI, launched as regular user, was dealing with the RPC model object just fine. However the model test (test/test_model.py) was failing in a test called test_asynctask for a reason that at first I didn't understand. After further investigation, the core reason was that this RPC has limitations about the python objects it can provide across the wire using serialization. The Model class does not contain elemental python structures only. It contains threads, for example. And the asynctask.py (invokd by test_asynctask AND uses threads) was not behaving as it should, causing non-trivial errors in test_model (thus wouldn't be usable by the UI as well). My attempts to try to fix it weren't promising. I've tried a serializer that is more generic to transmit the RPC, but this requires an update of PyRO version. The version that Fedora 19 installs by default is 4.12, I needed the latest (4.24). After the update I've been facing a lot of misbehavior from the RPC, things that were working weren't working anymore and so on. Then it occurred to me that this wouldn't be acceptable (or would be a huge pain) to maintain in all the distros kimchi supports, such as RHEL6.5. I do not believe we can rely on another 3rd party software behavior across all the supported distros for the proper behavior of kimchi. And there's also the uncertainty that even with the latest version, there's no guarantee that the model classes would get too complex to be "serializable" by the RPC and sooner of later stop working properly. That said, I'm putting this RPC effort to a halt. What I've started to do is try to launch a second cherrypy process, running as root, running only the backend classes (no UI enabled), to provide the frontend with the info it can't get from the regular user. I believe this approach will require more work, but it's certain to work across all distros without relying in another 3rd party software. As always, any help/tips are appreciated :) On 03/05/2014 11:49 AM, Daniel H Barboza wrote:
Information about the issue:
https://github.com/kimchi-project/kimchi/issues/329
We had some ideas about it in the last weekly scrum:
- creating an user 'kimchi' with a lot of privileges. It is the simplest of the solutions but implies in a lot of annoyances (different system paths between distros, libvirt does not work the same way in all distros). There is no tell about the amount of new bugs and issues that shall emerge from such change.
- Separating the UI and the backend. This is my personal favourite but I believe it is also the hardest. We can implement this by separating frontend and backend as 2 separate cherrypy processes, the backend runs as root and the frontend runs as a regular user. The communication would be done using the REST API. The other approach would be the backend running as a regular python daemon, with root privileges, and kimchi would communicate with it using RPC. I believe the latter is more elegant and the former is easier to implement.
Things to consider:
- Distro support: Ubuntu, for example, behaves very different from Fedora as far as libvirt is concerned. The packaging model (apt-get instead of yum) differs in support as well. I think it's a fair guess that RHEL 6 and Suse will have different behavior as well.
- VM visibility: in the first idea (different user) only the VMs created by this specific user would be visible to kimchi, unless we do something about it (tweaking libvirt configuration perhaps?).
- User authentication. Right now the user authentication presented in kimchi exists simply to authenticate it as a regular user of the host. The owner of the process will be root, doesn't matter which user logs in (and I guess this is the critical security flaw we have). Do we need ro rethink the authentication model as well?
Please provide your input and ideas!
Daniel
_______________________________________________ Kimchi-devel mailing list Kimchi-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/kimchi-devel
participants (4)
-
Aline Manera -
Daniel H Barboza -
Royce Lv -
Sheldon