Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Chris, I agree that Connect shouldn't obtusely refuse operator-helpful information on known, handled exceptions. The situation you describe here feels like a great example. In cases like this, I would suggest printing the message, and in the field that would contain the stack trace, replacing that message with the pre-defined string that says Connect is running in a mode that doesn't print stack traces. This way, the schema is not changed. In places where the message would include the stack trace along with information about the fault, omitting the stack trace or replacing it with the pre-defined string would work just fine for me. Thanks! Connor On 5/13/20, 12:30 PM, "Christopher Egerton" wrote: Hi Connor, I think this is really close but have one more thought. Uncaught exceptions in the REST API are different from exceptions that come about when tasks or connectors fail, and can be used for different purposes. Stack traces in 500 errors are probably only useful for the administrator of the Connect cluster. However, if a user has tried to create a connector and sees that it or one of its tasks has failed, a brief message about the cause of failure might actually be pretty helpful, and if they can't get any information on why the connector or task failed, then they're essentially at the mercy of the Connect cluster administrator for figuring out what caused the failure. Would it be alright to include the exception message, but not the entire stack trace, in the response for requests to view the status of a connector or task? Cheers, Chris On Wed, May 6, 2020 at 12:07 PM Connor Penhale wrote: > Hi Chris, > > Apologies for the name confusion! I've been working with the my customer > sponsor over the last few weeks, and we finally have an answer regarding > "only exceptions or all responses." This organization is really interested > in removing stack traces from all responses, which will expand the scope of > this KIP a bit. I'm going to update the wiki entry, and then would it be > reasonable to call for a vote? > > Thanks! > Connor > > On 4/17/20, 3:53 PM, "Christopher Egerton" wrote: > > Hi Connor, > > That's great, but I think you may have mistaken Colin for me :) > > One more thing that should be addressed--the "public interfaces" > section > isn't just for Java interfaces, it's for any changes to any public > part of > Kafka that users and external developers interact with. As far as > Connect > is concerned, this includes (but is not limited to) the REST API and > worker > configuration properties, so it might be worth briefly summarizing the > scope of your proposed changes in that section (something like "We > plan on > adding a new worker config named that will affect the REST API > under > ". > > Cheers, > > Chris > > On Wed, Apr 15, 2020 at 1:00 PM Connor Penhale > wrote: > > > Hi Chris, > > > > I can ask the customer if they can disclose any additional > information. I > > provided the information around "PCI-DSS" to give the community a > flavor of > > the type of environment the customer was operating in. The current > mode is > > /not/ insecure, I would agree with this. I would be willing to agree > that > > my customer has particular security audit requirements that go above > and > > beyond what most environments would consider reasonable. Are you > > comfortable with that language? > > > > " enable.rest.response.stack.traces" works great for me! > > > > I created a new class in the example PR because I wanted the highest > > chance of not gunking up the works by stepping on toes in an > important > > class. I figured I'd be reducing risk by creating an alternative > > implementing class. In retrospect, and now that I'm getting a > first-hand > > look at Kafka's community process, that is probably unnecessary. > > Additionally, I would agree with your statement that we should > modify the > > existing ExceptionMapper to avoid behavior divergence in subsequent > > releases and ensure this feature's particular scope is easy to > maintain. > > > > Thanks! > > Connor > > > > On 4/15/20, 1:17 PM, "Colin McCabe" wrote: > > > > Hi Connor, > > > > I still would like to hear more about whether this feature is > required > > for PCI-DSS or any other security certification. Nobody I talked to > seemed > > to think that it was-- if there are certifications that would > require this, > > it
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Connor, I think this is really close but have one more thought. Uncaught exceptions in the REST API are different from exceptions that come about when tasks or connectors fail, and can be used for different purposes. Stack traces in 500 errors are probably only useful for the administrator of the Connect cluster. However, if a user has tried to create a connector and sees that it or one of its tasks has failed, a brief message about the cause of failure might actually be pretty helpful, and if they can't get any information on why the connector or task failed, then they're essentially at the mercy of the Connect cluster administrator for figuring out what caused the failure. Would it be alright to include the exception message, but not the entire stack trace, in the response for requests to view the status of a connector or task? Cheers, Chris On Wed, May 6, 2020 at 12:07 PM Connor Penhale wrote: > Hi Chris, > > Apologies for the name confusion! I've been working with the my customer > sponsor over the last few weeks, and we finally have an answer regarding > "only exceptions or all responses." This organization is really interested > in removing stack traces from all responses, which will expand the scope of > this KIP a bit. I'm going to update the wiki entry, and then would it be > reasonable to call for a vote? > > Thanks! > Connor > > On 4/17/20, 3:53 PM, "Christopher Egerton" wrote: > > Hi Connor, > > That's great, but I think you may have mistaken Colin for me :) > > One more thing that should be addressed--the "public interfaces" > section > isn't just for Java interfaces, it's for any changes to any public > part of > Kafka that users and external developers interact with. As far as > Connect > is concerned, this includes (but is not limited to) the REST API and > worker > configuration properties, so it might be worth briefly summarizing the > scope of your proposed changes in that section (something like "We > plan on > adding a new worker config named that will affect the REST API > under > ". > > Cheers, > > Chris > > On Wed, Apr 15, 2020 at 1:00 PM Connor Penhale > wrote: > > > Hi Chris, > > > > I can ask the customer if they can disclose any additional > information. I > > provided the information around "PCI-DSS" to give the community a > flavor of > > the type of environment the customer was operating in. The current > mode is > > /not/ insecure, I would agree with this. I would be willing to agree > that > > my customer has particular security audit requirements that go above > and > > beyond what most environments would consider reasonable. Are you > > comfortable with that language? > > > > " enable.rest.response.stack.traces" works great for me! > > > > I created a new class in the example PR because I wanted the highest > > chance of not gunking up the works by stepping on toes in an > important > > class. I figured I'd be reducing risk by creating an alternative > > implementing class. In retrospect, and now that I'm getting a > first-hand > > look at Kafka's community process, that is probably unnecessary. > > Additionally, I would agree with your statement that we should > modify the > > existing ExceptionMapper to avoid behavior divergence in subsequent > > releases and ensure this feature's particular scope is easy to > maintain. > > > > Thanks! > > Connor > > > > On 4/15/20, 1:17 PM, "Colin McCabe" wrote: > > > > Hi Connor, > > > > I still would like to hear more about whether this feature is > required > > for PCI-DSS or any other security certification. Nobody I talked to > seemed > > to think that it was-- if there are certifications that would > require this, > > it would be nice to know. However, I don't object to implementing > this as > > long as we don't imply that the current mode is insecure. > > > > What do you think about using > "enable.rest.response.stack.traces" as > > the config name? It seems like that makes it clearer that it's a > boolean > > value. > > > > It's not really necessary to describe the internal > implementation in > > the KIP, but since you mentioned it, it's probably worth considering > using > > the current ExceptionMapper class with a different configuration > rather > > than creating a new one. > > > > best, > > Colin > > > > > > On Mon, Apr 13, 2020, at 09:04, Connor Penhale wrote: > > > Hi Chris! > > > > > > RE: SSL, indeed, the issue is not that the information is not > > > encrypted, but that there is no authorization layer. > > > > > > I'll be sure to edit the KIP as we continue discussion! > > > > > > RE: the 200 response you highlighted, great catch! I'll work > with my > > >
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Chris, Apologies for the name confusion! I've been working with the my customer sponsor over the last few weeks, and we finally have an answer regarding "only exceptions or all responses." This organization is really interested in removing stack traces from all responses, which will expand the scope of this KIP a bit. I'm going to update the wiki entry, and then would it be reasonable to call for a vote? Thanks! Connor On 4/17/20, 3:53 PM, "Christopher Egerton" wrote: Hi Connor, That's great, but I think you may have mistaken Colin for me :) One more thing that should be addressed--the "public interfaces" section isn't just for Java interfaces, it's for any changes to any public part of Kafka that users and external developers interact with. As far as Connect is concerned, this includes (but is not limited to) the REST API and worker configuration properties, so it might be worth briefly summarizing the scope of your proposed changes in that section (something like "We plan on adding a new worker config named that will affect the REST API under ". Cheers, Chris On Wed, Apr 15, 2020 at 1:00 PM Connor Penhale wrote: > Hi Chris, > > I can ask the customer if they can disclose any additional information. I > provided the information around "PCI-DSS" to give the community a flavor of > the type of environment the customer was operating in. The current mode is > /not/ insecure, I would agree with this. I would be willing to agree that > my customer has particular security audit requirements that go above and > beyond what most environments would consider reasonable. Are you > comfortable with that language? > > " enable.rest.response.stack.traces" works great for me! > > I created a new class in the example PR because I wanted the highest > chance of not gunking up the works by stepping on toes in an important > class. I figured I'd be reducing risk by creating an alternative > implementing class. In retrospect, and now that I'm getting a first-hand > look at Kafka's community process, that is probably unnecessary. > Additionally, I would agree with your statement that we should modify the > existing ExceptionMapper to avoid behavior divergence in subsequent > releases and ensure this feature's particular scope is easy to maintain. > > Thanks! > Connor > > On 4/15/20, 1:17 PM, "Colin McCabe" wrote: > > Hi Connor, > > I still would like to hear more about whether this feature is required > for PCI-DSS or any other security certification. Nobody I talked to seemed > to think that it was-- if there are certifications that would require this, > it would be nice to know. However, I don't object to implementing this as > long as we don't imply that the current mode is insecure. > > What do you think about using "enable.rest.response.stack.traces" as > the config name? It seems like that makes it clearer that it's a boolean > value. > > It's not really necessary to describe the internal implementation in > the KIP, but since you mentioned it, it's probably worth considering using > the current ExceptionMapper class with a different configuration rather > than creating a new one. > > best, > Colin > > > On Mon, Apr 13, 2020, at 09:04, Connor Penhale wrote: > > Hi Chris! > > > > RE: SSL, indeed, the issue is not that the information is not > > encrypted, but that there is no authorization layer. > > > > I'll be sure to edit the KIP as we continue discussion! > > > > RE: the 200 response you highlighted, great catch! I'll work with my > > customer and get back to you on their audit team's intention! I'm > > fairly certain I know the answer, but I need to be sure before I > speak > > for them. > > > > Thanks! > > Connor > > > > On 4/8/20, 11:27 PM, "Christopher Egerton" > wrote: > > > > Hi Connor, > > > > Just a few more remarks! > > > > I noticed that you said "Kafka Connect was passing these > exceptions without > > authentication." For what it's worth, the Connect REST API can > be secured > > with TLS out-of-the-box by configuring the worker with the > various ssl.* > > properties, but that doesn't provide any kind of authorization > layer to > > provide levels of security depending who the user is. Just > pointing out in > > case this helps with your use case. > > > > As far as editing the KIP based on discussion goes--it's not only > > acceptable, it's expected :) Ideally, the KIP should be kept > up-to-date
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Connor,
That's great, but I think you may have mistaken Colin for me :)
One more thing that should be addressed--the "public interfaces" section
isn't just for Java interfaces, it's for any changes to any public part of
Kafka that users and external developers interact with. As far as Connect
is concerned, this includes (but is not limited to) the REST API and worker
configuration properties, so it might be worth briefly summarizing the
scope of your proposed changes in that section (something like "We plan on
adding a new worker config named that will affect the REST API under
".
Cheers,
Chris
On Wed, Apr 15, 2020 at 1:00 PM Connor Penhale
wrote:
> Hi Chris,
>
> I can ask the customer if they can disclose any additional information. I
> provided the information around "PCI-DSS" to give the community a flavor of
> the type of environment the customer was operating in. The current mode is
> /not/ insecure, I would agree with this. I would be willing to agree that
> my customer has particular security audit requirements that go above and
> beyond what most environments would consider reasonable. Are you
> comfortable with that language?
>
> " enable.rest.response.stack.traces" works great for me!
>
> I created a new class in the example PR because I wanted the highest
> chance of not gunking up the works by stepping on toes in an important
> class. I figured I'd be reducing risk by creating an alternative
> implementing class. In retrospect, and now that I'm getting a first-hand
> look at Kafka's community process, that is probably unnecessary.
> Additionally, I would agree with your statement that we should modify the
> existing ExceptionMapper to avoid behavior divergence in subsequent
> releases and ensure this feature's particular scope is easy to maintain.
>
> Thanks!
> Connor
>
> On 4/15/20, 1:17 PM, "Colin McCabe" wrote:
>
> Hi Connor,
>
> I still would like to hear more about whether this feature is required
> for PCI-DSS or any other security certification. Nobody I talked to seemed
> to think that it was-- if there are certifications that would require this,
> it would be nice to know. However, I don't object to implementing this as
> long as we don't imply that the current mode is insecure.
>
> What do you think about using "enable.rest.response.stack.traces" as
> the config name? It seems like that makes it clearer that it's a boolean
> value.
>
> It's not really necessary to describe the internal implementation in
> the KIP, but since you mentioned it, it's probably worth considering using
> the current ExceptionMapper class with a different configuration rather
> than creating a new one.
>
> best,
> Colin
>
>
> On Mon, Apr 13, 2020, at 09:04, Connor Penhale wrote:
> > Hi Chris!
> >
> > RE: SSL, indeed, the issue is not that the information is not
> > encrypted, but that there is no authorization layer.
> >
> > I'll be sure to edit the KIP as we continue discussion!
> >
> > RE: the 200 response you highlighted, great catch! I'll work with my
> > customer and get back to you on their audit team's intention! I'm
> > fairly certain I know the answer, but I need to be sure before I
> speak
> > for them.
> >
> > Thanks!
> > Connor
> >
> > On 4/8/20, 11:27 PM, "Christopher Egerton"
> wrote:
> >
> > Hi Connor,
> >
> > Just a few more remarks!
> >
> > I noticed that you said "Kafka Connect was passing these
> exceptions without
> > authentication." For what it's worth, the Connect REST API can
> be secured
> > with TLS out-of-the-box by configuring the worker with the
> various ssl.*
> > properties, but that doesn't provide any kind of authorization
> layer to
> > provide levels of security depending who the user is. Just
> pointing out in
> > case this helps with your use case.
> >
> > As far as editing the KIP based on discussion goes--it's not only
> > acceptable, it's expected :) Ideally, the KIP should be kept
> up-to-date to
> > the point where, were it to be accepted at any moment, it would
> accurately
> > reflect the changes that would then be made to Kafka. This can
> be relaxed
> > if there's rapid iteration or items that are still up for
> discussion, but
> > as soon as things settle down it should be updated.
> >
> > As far as item 4 goes, my question was about exceptions that
> aren't handled
> > by the ExceptionMapper, but which are returned as part of the
> response body
> > when querying the status of a connector or task that has failed
> by querying
> > the /connectors/{name}/status or
> /connectors/{name}/tasks/{taskId}/status
> > endpoints. Even if the request is successful and results in an
> HTTP 200
> > response, the body might contain a stack trace if the connector
> or any of
> > its
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Chris,
I can ask the customer if they can disclose any additional information. I
provided the information around "PCI-DSS" to give the community a flavor of the
type of environment the customer was operating in. The current mode is /not/
insecure, I would agree with this. I would be willing to agree that my customer
has particular security audit requirements that go above and beyond what most
environments would consider reasonable. Are you comfortable with that language?
" enable.rest.response.stack.traces" works great for me!
I created a new class in the example PR because I wanted the highest chance of
not gunking up the works by stepping on toes in an important class. I figured
I'd be reducing risk by creating an alternative implementing class. In
retrospect, and now that I'm getting a first-hand look at Kafka's community
process, that is probably unnecessary. Additionally, I would agree with your
statement that we should modify the existing ExceptionMapper to avoid behavior
divergence in subsequent releases and ensure this feature's particular scope is
easy to maintain.
Thanks!
Connor
On 4/15/20, 1:17 PM, "Colin McCabe" wrote:
Hi Connor,
I still would like to hear more about whether this feature is required for
PCI-DSS or any other security certification. Nobody I talked to seemed to
think that it was-- if there are certifications that would require this, it
would be nice to know. However, I don't object to implementing this as long as
we don't imply that the current mode is insecure.
What do you think about using "enable.rest.response.stack.traces" as the
config name? It seems like that makes it clearer that it's a boolean value.
It's not really necessary to describe the internal implementation in the
KIP, but since you mentioned it, it's probably worth considering using the
current ExceptionMapper class with a different configuration rather than
creating a new one.
best,
Colin
On Mon, Apr 13, 2020, at 09:04, Connor Penhale wrote:
> Hi Chris!
>
> RE: SSL, indeed, the issue is not that the information is not
> encrypted, but that there is no authorization layer.
>
> I'll be sure to edit the KIP as we continue discussion!
>
> RE: the 200 response you highlighted, great catch! I'll work with my
> customer and get back to you on their audit team's intention! I'm
> fairly certain I know the answer, but I need to be sure before I speak
> for them.
>
> Thanks!
> Connor
>
> On 4/8/20, 11:27 PM, "Christopher Egerton" wrote:
>
> Hi Connor,
>
> Just a few more remarks!
>
> I noticed that you said "Kafka Connect was passing these exceptions
without
> authentication." For what it's worth, the Connect REST API can be
secured
> with TLS out-of-the-box by configuring the worker with the various
ssl.*
> properties, but that doesn't provide any kind of authorization layer
to
> provide levels of security depending who the user is. Just pointing
out in
> case this helps with your use case.
>
> As far as editing the KIP based on discussion goes--it's not only
> acceptable, it's expected :) Ideally, the KIP should be kept
up-to-date to
> the point where, were it to be accepted at any moment, it would
accurately
> reflect the changes that would then be made to Kafka. This can be
relaxed
> if there's rapid iteration or items that are still up for discussion,
but
> as soon as things settle down it should be updated.
>
> As far as item 4 goes, my question was about exceptions that aren't
handled
> by the ExceptionMapper, but which are returned as part of the
response body
> when querying the status of a connector or task that has failed by
querying
> the /connectors/{name}/status or
/connectors/{name}/tasks/{taskId}/status
> endpoints. Even if the request is successful and results in an HTTP
200
> response, the body might contain a stack trace if the connector or
any of
> its tasks have failed.
>
> For example, I ran an instance of the FileStreamSource connector named
> "file-source" locally and instructed it to consume from a file that it
> lacked permissions to read. When I queried the status of that
connector by
> issuing a request to /connectors/file-source/status, I got back the
> following response:
>
> {
> "name": "file-source",
> "connector": {
> "state": "RUNNING",
> "worker_id": "192.168.86.21:8083"
> },
> "tasks": [
> {
> "id": 0,
> "state": "FAILED",
> "worker_id": "192.168.86.21:8083",
> "trace": "org.apache.kafka.connect.errors.ConnectException:
> java.nio.file.AccessDeniedException:
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Connor,
I still would like to hear more about whether this feature is required for
PCI-DSS or any other security certification. Nobody I talked to seemed to
think that it was-- if there are certifications that would require this, it
would be nice to know. However, I don't object to implementing this as long as
we don't imply that the current mode is insecure.
What do you think about using "enable.rest.response.stack.traces" as the config
name? It seems like that makes it clearer that it's a boolean value.
It's not really necessary to describe the internal implementation in the KIP,
but since you mentioned it, it's probably worth considering using the current
ExceptionMapper class with a different configuration rather than creating a new
one.
best,
Colin
On Mon, Apr 13, 2020, at 09:04, Connor Penhale wrote:
> Hi Chris!
>
> RE: SSL, indeed, the issue is not that the information is not
> encrypted, but that there is no authorization layer.
>
> I'll be sure to edit the KIP as we continue discussion!
>
> RE: the 200 response you highlighted, great catch! I'll work with my
> customer and get back to you on their audit team's intention! I'm
> fairly certain I know the answer, but I need to be sure before I speak
> for them.
>
> Thanks!
> Connor
>
> On 4/8/20, 11:27 PM, "Christopher Egerton" wrote:
>
> Hi Connor,
>
> Just a few more remarks!
>
> I noticed that you said "Kafka Connect was passing these exceptions
> without
> authentication." For what it's worth, the Connect REST API can be secured
> with TLS out-of-the-box by configuring the worker with the various ssl.*
> properties, but that doesn't provide any kind of authorization layer to
> provide levels of security depending who the user is. Just pointing out in
> case this helps with your use case.
>
> As far as editing the KIP based on discussion goes--it's not only
> acceptable, it's expected :) Ideally, the KIP should be kept up-to-date to
> the point where, were it to be accepted at any moment, it would accurately
> reflect the changes that would then be made to Kafka. This can be relaxed
> if there's rapid iteration or items that are still up for discussion, but
> as soon as things settle down it should be updated.
>
> As far as item 4 goes, my question was about exceptions that aren't
> handled
> by the ExceptionMapper, but which are returned as part of the response
> body
> when querying the status of a connector or task that has failed by
> querying
> the /connectors/{name}/status or /connectors/{name}/tasks/{taskId}/status
> endpoints. Even if the request is successful and results in an HTTP 200
> response, the body might contain a stack trace if the connector or any of
> its tasks have failed.
>
> For example, I ran an instance of the FileStreamSource connector named
> "file-source" locally and instructed it to consume from a file that it
> lacked permissions to read. When I queried the status of that connector by
> issuing a request to /connectors/file-source/status, I got back the
> following response:
>
> {
> "name": "file-source",
> "connector": {
> "state": "RUNNING",
> "worker_id": "192.168.86.21:8083"
> },
> "tasks": [
> {
> "id": 0,
> "state": "FAILED",
> "worker_id": "192.168.86.21:8083",
> "trace": "org.apache.kafka.connect.errors.ConnectException:
> java.nio.file.AccessDeniedException: test.txt\n\tat
>
> org.apache.kafka.connect.file.FileStreamSourceTask.poll(FileStreamSourceTask.java:116)\n\tat
>
> org.apache.kafka.connect.runtime.WorkerSourceTask.poll(WorkerSourceTask.java:265)\n\tat
>
> org.apache.kafka.connect.runtime.WorkerSourceTask.execute(WorkerSourceTask.java:232)\n\tat
>
> org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:177)\n\tat
>
> org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:227)\n\tat
>
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)\n\tat
> java.util.concurrent.FutureTask.run(FutureTask.java:266)\n\tat
>
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)\n\tat
>
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)\n\tat
> java.lang.Thread.run(Thread.java:748)\nCaused by:
> java.nio.file.AccessDeniedException: test.txt\n\tat
>
> sun.nio.fs.UnixException.translateToIOException(UnixException.java:84)\n\tat
>
> sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:102)\n\tat
>
> sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:107)\n\tat
>
> sun.nio.fs.UnixFileSystemProvider.newByteChannel(UnixFileSystemProvider.java:214)\n\tat
> java.nio.file.Files.newByteChannel(Files.java:361)\n\tat
> java.nio.file.Files.newByteChannel(Files.java:407)\n\tat
>
>
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Chris!
RE: SSL, indeed, the issue is not that the information is not encrypted, but
that there is no authorization layer.
I'll be sure to edit the KIP as we continue discussion!
RE: the 200 response you highlighted, great catch! I'll work with my customer
and get back to you on their audit team's intention! I'm fairly certain I know
the answer, but I need to be sure before I speak for them.
Thanks!
Connor
On 4/8/20, 11:27 PM, "Christopher Egerton" wrote:
Hi Connor,
Just a few more remarks!
I noticed that you said "Kafka Connect was passing these exceptions without
authentication." For what it's worth, the Connect REST API can be secured
with TLS out-of-the-box by configuring the worker with the various ssl.*
properties, but that doesn't provide any kind of authorization layer to
provide levels of security depending who the user is. Just pointing out in
case this helps with your use case.
As far as editing the KIP based on discussion goes--it's not only
acceptable, it's expected :) Ideally, the KIP should be kept up-to-date to
the point where, were it to be accepted at any moment, it would accurately
reflect the changes that would then be made to Kafka. This can be relaxed
if there's rapid iteration or items that are still up for discussion, but
as soon as things settle down it should be updated.
As far as item 4 goes, my question was about exceptions that aren't handled
by the ExceptionMapper, but which are returned as part of the response body
when querying the status of a connector or task that has failed by querying
the /connectors/{name}/status or /connectors/{name}/tasks/{taskId}/status
endpoints. Even if the request is successful and results in an HTTP 200
response, the body might contain a stack trace if the connector or any of
its tasks have failed.
For example, I ran an instance of the FileStreamSource connector named
"file-source" locally and instructed it to consume from a file that it
lacked permissions to read. When I queried the status of that connector by
issuing a request to /connectors/file-source/status, I got back the
following response:
{
"name": "file-source",
"connector": {
"state": "RUNNING",
"worker_id": "192.168.86.21:8083"
},
"tasks": [
{
"id": 0,
"state": "FAILED",
"worker_id": "192.168.86.21:8083",
"trace": "org.apache.kafka.connect.errors.ConnectException:
java.nio.file.AccessDeniedException: test.txt\n\tat
org.apache.kafka.connect.file.FileStreamSourceTask.poll(FileStreamSourceTask.java:116)\n\tat
org.apache.kafka.connect.runtime.WorkerSourceTask.poll(WorkerSourceTask.java:265)\n\tat
org.apache.kafka.connect.runtime.WorkerSourceTask.execute(WorkerSourceTask.java:232)\n\tat
org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:177)\n\tat
org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:227)\n\tat
java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)\n\tat
java.util.concurrent.FutureTask.run(FutureTask.java:266)\n\tat
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)\n\tat
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)\n\tat
java.lang.Thread.run(Thread.java:748)\nCaused by:
java.nio.file.AccessDeniedException: test.txt\n\tat
sun.nio.fs.UnixException.translateToIOException(UnixException.java:84)\n\tat
sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:102)\n\tat
sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:107)\n\tat
sun.nio.fs.UnixFileSystemProvider.newByteChannel(UnixFileSystemProvider.java:214)\n\tat
java.nio.file.Files.newByteChannel(Files.java:361)\n\tat
java.nio.file.Files.newByteChannel(Files.java:407)\n\tat
java.nio.file.spi.FileSystemProvider.newInputStream(FileSystemProvider.java:384)\n\tat
java.nio.file.Files.newInputStream(Files.java:152)\n\tat
org.apache.kafka.connect.file.FileStreamSourceTask.poll(FileStreamSourceTask.java:82)\n\t...
9 more\n"
}
],
"type": "source"
}
Note the "trace" field in the first element of the "tasks" field of the
response: this was the stack trace for the exception that caused the task
to fail during execution, which has nothing to do with the success or
failure of the REST request I issued to the /connectors/file-source/status
endpoint.
I was wondering if you wanted to include these kinds of stack traces as
part of the KIP, as opposed to uncaught exceptions that result in a 500
error from the REST API.
Cheers,
Chris
On Wed, Apr 8, 2020 at 9:51 AM Connor Penhale wrote:
> Hi All!
>
> Is there any additional feedback that the community can provide me on the
> KIP? Has anyone else run into requirements like
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Connor,
Just a few more remarks!
I noticed that you said "Kafka Connect was passing these exceptions without
authentication." For what it's worth, the Connect REST API can be secured
with TLS out-of-the-box by configuring the worker with the various ssl.*
properties, but that doesn't provide any kind of authorization layer to
provide levels of security depending who the user is. Just pointing out in
case this helps with your use case.
As far as editing the KIP based on discussion goes--it's not only
acceptable, it's expected :) Ideally, the KIP should be kept up-to-date to
the point where, were it to be accepted at any moment, it would accurately
reflect the changes that would then be made to Kafka. This can be relaxed
if there's rapid iteration or items that are still up for discussion, but
as soon as things settle down it should be updated.
As far as item 4 goes, my question was about exceptions that aren't handled
by the ExceptionMapper, but which are returned as part of the response body
when querying the status of a connector or task that has failed by querying
the /connectors/{name}/status or /connectors/{name}/tasks/{taskId}/status
endpoints. Even if the request is successful and results in an HTTP 200
response, the body might contain a stack trace if the connector or any of
its tasks have failed.
For example, I ran an instance of the FileStreamSource connector named
"file-source" locally and instructed it to consume from a file that it
lacked permissions to read. When I queried the status of that connector by
issuing a request to /connectors/file-source/status, I got back the
following response:
{
"name": "file-source",
"connector": {
"state": "RUNNING",
"worker_id": "192.168.86.21:8083"
},
"tasks": [
{
"id": 0,
"state": "FAILED",
"worker_id": "192.168.86.21:8083",
"trace": "org.apache.kafka.connect.errors.ConnectException:
java.nio.file.AccessDeniedException: test.txt\n\tat
org.apache.kafka.connect.file.FileStreamSourceTask.poll(FileStreamSourceTask.java:116)\n\tat
org.apache.kafka.connect.runtime.WorkerSourceTask.poll(WorkerSourceTask.java:265)\n\tat
org.apache.kafka.connect.runtime.WorkerSourceTask.execute(WorkerSourceTask.java:232)\n\tat
org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:177)\n\tat
org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:227)\n\tat
java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)\n\tat
java.util.concurrent.FutureTask.run(FutureTask.java:266)\n\tat
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)\n\tat
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)\n\tat
java.lang.Thread.run(Thread.java:748)\nCaused by:
java.nio.file.AccessDeniedException: test.txt\n\tat
sun.nio.fs.UnixException.translateToIOException(UnixException.java:84)\n\tat
sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:102)\n\tat
sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:107)\n\tat
sun.nio.fs.UnixFileSystemProvider.newByteChannel(UnixFileSystemProvider.java:214)\n\tat
java.nio.file.Files.newByteChannel(Files.java:361)\n\tat
java.nio.file.Files.newByteChannel(Files.java:407)\n\tat
java.nio.file.spi.FileSystemProvider.newInputStream(FileSystemProvider.java:384)\n\tat
java.nio.file.Files.newInputStream(Files.java:152)\n\tat
org.apache.kafka.connect.file.FileStreamSourceTask.poll(FileStreamSourceTask.java:82)\n\t...
9 more\n"
}
],
"type": "source"
}
Note the "trace" field in the first element of the "tasks" field of the
response: this was the stack trace for the exception that caused the task
to fail during execution, which has nothing to do with the success or
failure of the REST request I issued to the /connectors/file-source/status
endpoint.
I was wondering if you wanted to include these kinds of stack traces as
part of the KIP, as opposed to uncaught exceptions that result in a 500
error from the REST API.
Cheers,
Chris
On Wed, Apr 8, 2020 at 9:51 AM Connor Penhale wrote:
> Hi All!
>
> Is there any additional feedback that the community can provide me on the
> KIP? Has anyone else run into requirements like this, or maybe my customer
> is the only one :)? If the scope looks good, is it time to call a vote? Or
> should I start porting my 2.0 code to 2.6 to show examples?
>
> Thanks!
> Connor
>
> On 4/6/20, 9:03 AM, "Connor Penhale" wrote:
>
> Hi Colin,
>
> We did not find a specific security vulnerability. Our customer had
> auditors in their environment, and they identified Kafka Connect as out of
> compliance with their particular standards, something that happens all the
> time for REST-based applications. What these security auditors expected
> Kafka Connect to be able to do was tune the response. As Kafka Connect
> could not provide this functionality, I'm proposing this KIP. Does that
> make sense? Should I still go through the process of a security disclosure?
>
> Our
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi All! Is there any additional feedback that the community can provide me on the KIP? Has anyone else run into requirements like this, or maybe my customer is the only one :)? If the scope looks good, is it time to call a vote? Or should I start porting my 2.0 code to 2.6 to show examples? Thanks! Connor On 4/6/20, 9:03 AM, "Connor Penhale" wrote: Hi Colin, We did not find a specific security vulnerability. Our customer had auditors in their environment, and they identified Kafka Connect as out of compliance with their particular standards, something that happens all the time for REST-based applications. What these security auditors expected Kafka Connect to be able to do was tune the response. As Kafka Connect could not provide this functionality, I'm proposing this KIP. Does that make sense? Should I still go through the process of a security disclosure? Our particular need was around suppressing exceptions in the "public" response, as Kafka Connect was passing these exceptions without authentication, they became a public endpoint upon which the auditors could fuzz, and show it being out of compliance. Keeping these exceptions in the logs, as proposed in the KIP, makes sense to me as an operator. I only mention PCI-DSS as this was the kind of environment my customer had that was making the request for being able to tune the response. Thanks! Connor --- Connor Penhale | Enterprise Architect, OpenLogic (https://openlogic.com/) Perforce (https://www.perforce.com/) Support: +1 866.399.6736 On 4/3/20, 3:24 PM, "Colin McCabe" wrote: Also, if you do find a security issue, the process to follow is here: https://kafka.apache.org/project-security.html . best, Colin On Fri, Apr 3, 2020, at 14:20, Colin McCabe wrote: > Hi Connor, > > If we are putting security-sensitive information into REST responses, > that is a bug that needs to be fixed, not worked around with a > configuration option. Do you have an example of security-sensitive > information appearing in the exception text? Why do you feel that > PCI-DSS requires this change? > > By the way, the same concern applies to log messages. We do not log > sensitive information such as passwords to the log4j output. If you > know of that happening somewhere, please file a bug so it can be fixed. > > best, > Colin > > > On Fri, Apr 3, 2020, at 12:56, Connor Penhale wrote: > > Hi Chris! > > > > Thanks for your feedback! I'll number my responses to your questions / thoughts. > > > > 1. Apologies on that lack of clarity! I settled on "Detailed exception > > information has been suppressed. Please see logs." > > (https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R34). Should I update the KIP to reflect what I've already thought about? It's my first one, not sure what the process should be for editing. > > > > 2. I was unaware of the REST extensions! I'll see if I can implement > > the same behavior as a REST extension. I agree that the KIP still has > > merit, regardless of the feasibility of the extension, but in regards > > to the 5th thought, this might make that decision easier. > > > > 3. I agree with your suggestion here. Absolutely ready to take the > > community feedback on what makes sense here. > > > > 4. I should note that while I emphasized uncaught exceptions, I mean > > all exceptions handled by the ExceptionMapper, including > > ConnectRestExceptions. An example of this is here: > > https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R46 > > > > 5. I didn't know how specific I should get if I had already taken a > > stab at implementing! I'm happy to edit this in whatever way we want to > > go about it. > > > > Let me know if anyone has any other questions or feedback! > > > > > > Thanks! > > Connor > > > > On 4/2/20, 3:58 PM, "Christopher Egerton" wrote: > > > > Hi Connor, > > > > Great stuff! I generally like being able to see the stack trace of an > > exception directly via the REST API but can definitely understand the > > security concerns here. I've got a few questions/remarks about the KIP and > > would be interested in your thoughts: > > > > 1. The KIP mentions a SUPRESSED_EXCEPTION_MESSAGE, but doesn't actually > > outline what this message would actually be. It'd be great to see the > > actual message in the KIP since people may have thoughts on
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Colin, We did not find a specific security vulnerability. Our customer had auditors in their environment, and they identified Kafka Connect as out of compliance with their particular standards, something that happens all the time for REST-based applications. What these security auditors expected Kafka Connect to be able to do was tune the response. As Kafka Connect could not provide this functionality, I'm proposing this KIP. Does that make sense? Should I still go through the process of a security disclosure? Our particular need was around suppressing exceptions in the "public" response, as Kafka Connect was passing these exceptions without authentication, they became a public endpoint upon which the auditors could fuzz, and show it being out of compliance. Keeping these exceptions in the logs, as proposed in the KIP, makes sense to me as an operator. I only mention PCI-DSS as this was the kind of environment my customer had that was making the request for being able to tune the response. Thanks! Connor --- Connor Penhale | Enterprise Architect, OpenLogic (https://openlogic.com/) Perforce (https://www.perforce.com/) Support: +1 866.399.6736 On 4/3/20, 3:24 PM, "Colin McCabe" wrote: Also, if you do find a security issue, the process to follow is here: https://kafka.apache.org/project-security.html . best, Colin On Fri, Apr 3, 2020, at 14:20, Colin McCabe wrote: > Hi Connor, > > If we are putting security-sensitive information into REST responses, > that is a bug that needs to be fixed, not worked around with a > configuration option. Do you have an example of security-sensitive > information appearing in the exception text? Why do you feel that > PCI-DSS requires this change? > > By the way, the same concern applies to log messages. We do not log > sensitive information such as passwords to the log4j output. If you > know of that happening somewhere, please file a bug so it can be fixed. > > best, > Colin > > > On Fri, Apr 3, 2020, at 12:56, Connor Penhale wrote: > > Hi Chris! > > > > Thanks for your feedback! I'll number my responses to your questions / thoughts. > > > > 1. Apologies on that lack of clarity! I settled on "Detailed exception > > information has been suppressed. Please see logs." > > (https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R34). Should I update the KIP to reflect what I've already thought about? It's my first one, not sure what the process should be for editing. > > > > 2. I was unaware of the REST extensions! I'll see if I can implement > > the same behavior as a REST extension. I agree that the KIP still has > > merit, regardless of the feasibility of the extension, but in regards > > to the 5th thought, this might make that decision easier. > > > > 3. I agree with your suggestion here. Absolutely ready to take the > > community feedback on what makes sense here. > > > > 4. I should note that while I emphasized uncaught exceptions, I mean > > all exceptions handled by the ExceptionMapper, including > > ConnectRestExceptions. An example of this is here: > > https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R46 > > > > 5. I didn't know how specific I should get if I had already taken a > > stab at implementing! I'm happy to edit this in whatever way we want to > > go about it. > > > > Let me know if anyone has any other questions or feedback! > > > > > > Thanks! > > Connor > > > > On 4/2/20, 3:58 PM, "Christopher Egerton" wrote: > > > > Hi Connor, > > > > Great stuff! I generally like being able to see the stack trace of an > > exception directly via the REST API but can definitely understand the > > security concerns here. I've got a few questions/remarks about the KIP and > > would be interested in your thoughts: > > > > 1. The KIP mentions a SUPRESSED_EXCEPTION_MESSAGE, but doesn't actually > > outline what this message would actually be. It'd be great to see the > > actual message in the KIP since people may have thoughts on what it should > > be and want to comment on it as part of this discussion. > > > > 2. In the "Rejected Alternatives" section, an Nginx proxy is > > mentioned as > > one possible way to filter out stack traces from the REST API. It > > seems > > like a Connect REST extension ( > > > > https://kafka.apache.org/24/javadoc/index.html?org/apache/kafka/connect/rest/ConnectRestExtension.html) > > would be a better alternative than an Nginx proxy; had you > > considered > > utilizing one? I still think this KIP is worthwhile and a REST > > extension > > shouldn't be necessary in
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Also, if you do find a security issue, the process to follow is here: https://kafka.apache.org/project-security.html . best, Colin On Fri, Apr 3, 2020, at 14:20, Colin McCabe wrote: > Hi Connor, > > If we are putting security-sensitive information into REST responses, > that is a bug that needs to be fixed, not worked around with a > configuration option. Do you have an example of security-sensitive > information appearing in the exception text? Why do you feel that > PCI-DSS requires this change? > > By the way, the same concern applies to log messages. We do not log > sensitive information such as passwords to the log4j output. If you > know of that happening somewhere, please file a bug so it can be fixed. > > best, > Colin > > > On Fri, Apr 3, 2020, at 12:56, Connor Penhale wrote: > > Hi Chris! > > > > Thanks for your feedback! I'll number my responses to your questions / > > thoughts. > > > > 1. Apologies on that lack of clarity! I settled on "Detailed exception > > information has been suppressed. Please see logs." > > (https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R34). > > Should I update the KIP to reflect what I've already thought about? It's > > my first one, not sure what the process should be for editing. > > > > 2. I was unaware of the REST extensions! I'll see if I can implement > > the same behavior as a REST extension. I agree that the KIP still has > > merit, regardless of the feasibility of the extension, but in regards > > to the 5th thought, this might make that decision easier. > > > > 3. I agree with your suggestion here. Absolutely ready to take the > > community feedback on what makes sense here. > > > > 4. I should note that while I emphasized uncaught exceptions, I mean > > all exceptions handled by the ExceptionMapper, including > > ConnectRestExceptions. An example of this is here: > > https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R46 > > > > 5. I didn't know how specific I should get if I had already taken a > > stab at implementing! I'm happy to edit this in whatever way we want to > > go about it. > > > > Let me know if anyone has any other questions or feedback! > > > > > > Thanks! > > Connor > > > > On 4/2/20, 3:58 PM, "Christopher Egerton" wrote: > > > > Hi Connor, > > > > Great stuff! I generally like being able to see the stack trace of an > > exception directly via the REST API but can definitely understand the > > security concerns here. I've got a few questions/remarks about the KIP > > and > > would be interested in your thoughts: > > > > 1. The KIP mentions a SUPRESSED_EXCEPTION_MESSAGE, but doesn't actually > > outline what this message would actually be. It'd be great to see the > > actual message in the KIP since people may have thoughts on what it > > should > > be and want to comment on it as part of this discussion. > > > > 2. In the "Rejected Alternatives" section, an Nginx proxy is > > mentioned as > > one possible way to filter out stack traces from the REST API. It > > seems > > like a Connect REST extension ( > > > > https://kafka.apache.org/24/javadoc/index.html?org/apache/kafka/connect/rest/ConnectRestExtension.html) > > would be a better alternative than an Nginx proxy; had you > > considered > > utilizing one? I still think this KIP is worthwhile and a REST > > extension > > shouldn't be necessary in order to lock down the REST API this way, > > but it > > might be worth calling out as an alternative and perhaps even a > > workaround > > in cases where users are stuck on a given version of Connect and > > can't > > upgrade to 2.6 (or whichever version this KIP lands on) any time > > soon. > > > > 3. The "error.rest.response.message.detail.enabled" property is a bit > > of a > > mouthful; it'd be great if we could come up with something more > > succinct. > > What do you think about something like "rest.response.stack.traces"? > > > > 4. The KIP is targeted at stack traces for uncaught exceptions, but it's > > also possible that stack traces get exposed in the REST API when > > querying > > the status of a connector or one of its tasks. Was this intentional? If > > so, > > it'd be great to call out why that kind of filtering is not required in > > the > > "Rejected Alternatives" section, and if not, it's probably not too late > > to > > consider modifying the KIP to cover those cases as well. > > > > 5. The KIP mentions creating a new, separate exception mapper class. > > This > > seems like more of an implementation detail and something that can be > > decided on during code review; unless it's critical to the functionality > > that the KIP aims to accomplish, I'd suggest leaving that part out > > since it > > shouldn't affect the impact on users of the Connect framework. >
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Connor, If we are putting security-sensitive information into REST responses, that is a bug that needs to be fixed, not worked around with a configuration option. Do you have an example of security-sensitive information appearing in the exception text? Why do you feel that PCI-DSS requires this change? By the way, the same concern applies to log messages. We do not log sensitive information such as passwords to the log4j output. If you know of that happening somewhere, please file a bug so it can be fixed. best, Colin On Fri, Apr 3, 2020, at 12:56, Connor Penhale wrote: > Hi Chris! > > Thanks for your feedback! I'll number my responses to your questions / > thoughts. > > 1. Apologies on that lack of clarity! I settled on "Detailed exception > information has been suppressed. Please see logs." > (https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R34). > Should I update the KIP to reflect what I've already thought about? It's my > first one, not sure what the process should be for editing. > > 2. I was unaware of the REST extensions! I'll see if I can implement > the same behavior as a REST extension. I agree that the KIP still has > merit, regardless of the feasibility of the extension, but in regards > to the 5th thought, this might make that decision easier. > > 3. I agree with your suggestion here. Absolutely ready to take the > community feedback on what makes sense here. > > 4. I should note that while I emphasized uncaught exceptions, I mean > all exceptions handled by the ExceptionMapper, including > ConnectRestExceptions. An example of this is here: > https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R46 > > 5. I didn't know how specific I should get if I had already taken a > stab at implementing! I'm happy to edit this in whatever way we want to > go about it. > > Let me know if anyone has any other questions or feedback! > > > Thanks! > Connor > > On 4/2/20, 3:58 PM, "Christopher Egerton" wrote: > > Hi Connor, > > Great stuff! I generally like being able to see the stack trace of an > exception directly via the REST API but can definitely understand the > security concerns here. I've got a few questions/remarks about the KIP and > would be interested in your thoughts: > > 1. The KIP mentions a SUPRESSED_EXCEPTION_MESSAGE, but doesn't actually > outline what this message would actually be. It'd be great to see the > actual message in the KIP since people may have thoughts on what it should > be and want to comment on it as part of this discussion. > > 2. In the "Rejected Alternatives" section, an Nginx proxy is > mentioned as > one possible way to filter out stack traces from the REST API. It > seems > like a Connect REST extension ( > > https://kafka.apache.org/24/javadoc/index.html?org/apache/kafka/connect/rest/ConnectRestExtension.html) > would be a better alternative than an Nginx proxy; had you > considered > utilizing one? I still think this KIP is worthwhile and a REST > extension > shouldn't be necessary in order to lock down the REST API this way, > but it > might be worth calling out as an alternative and perhaps even a > workaround > in cases where users are stuck on a given version of Connect and > can't > upgrade to 2.6 (or whichever version this KIP lands on) any time > soon. > > 3. The "error.rest.response.message.detail.enabled" property is a bit of a > mouthful; it'd be great if we could come up with something more succinct. > What do you think about something like "rest.response.stack.traces"? > > 4. The KIP is targeted at stack traces for uncaught exceptions, but it's > also possible that stack traces get exposed in the REST API when querying > the status of a connector or one of its tasks. Was this intentional? If > so, > it'd be great to call out why that kind of filtering is not required in > the > "Rejected Alternatives" section, and if not, it's probably not too late to > consider modifying the KIP to cover those cases as well. > > 5. The KIP mentions creating a new, separate exception mapper class. This > seems like more of an implementation detail and something that can be > decided on during code review; unless it's critical to the functionality > that the KIP aims to accomplish, I'd suggest leaving that part out since > it > shouldn't affect the impact on users of the Connect framework. > > Thanks for the KIP, looking forward to seeing this happen! > > Cheers, > > Chris > > On Thu, Apr 2, 2020 at 11:01 AM Connor Penhale > wrote: > > > Hello All! > > > > I’ve created the following KIP: > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-587:+Suppress+detailed+responses+for+handled+exceptions+in+security-sensitive+environments > > > > The PR that
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Chris! Thanks for your feedback! I'll number my responses to your questions / thoughts. 1. Apologies on that lack of clarity! I settled on "Detailed exception information has been suppressed. Please see logs." (https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R34). Should I update the KIP to reflect what I've already thought about? It's my first one, not sure what the process should be for editing. 2. I was unaware of the REST extensions! I'll see if I can implement the same behavior as a REST extension. I agree that the KIP still has merit, regardless of the feasibility of the extension, but in regards to the 5th thought, this might make that decision easier. 3. I agree with your suggestion here. Absolutely ready to take the community feedback on what makes sense here. 4. I should note that while I emphasized uncaught exceptions, I mean all exceptions handled by the ExceptionMapper, including ConnectRestExceptions. An example of this is here: https://github.com/apache/kafka/pull/8355/files#diff-64c265986e7bbe40cdd79f831e961907R46 5. I didn't know how specific I should get if I had already taken a stab at implementing! I'm happy to edit this in whatever way we want to go about it. Let me know if anyone has any other questions or feedback! Thanks! Connor On 4/2/20, 3:58 PM, "Christopher Egerton" wrote: Hi Connor, Great stuff! I generally like being able to see the stack trace of an exception directly via the REST API but can definitely understand the security concerns here. I've got a few questions/remarks about the KIP and would be interested in your thoughts: 1. The KIP mentions a SUPRESSED_EXCEPTION_MESSAGE, but doesn't actually outline what this message would actually be. It'd be great to see the actual message in the KIP since people may have thoughts on what it should be and want to comment on it as part of this discussion. 2. In the "Rejected Alternatives" section, an Nginx proxy is mentioned as one possible way to filter out stack traces from the REST API. It seems like a Connect REST extension ( https://kafka.apache.org/24/javadoc/index.html?org/apache/kafka/connect/rest/ConnectRestExtension.html) would be a better alternative than an Nginx proxy; had you considered utilizing one? I still think this KIP is worthwhile and a REST extension shouldn't be necessary in order to lock down the REST API this way, but it might be worth calling out as an alternative and perhaps even a workaround in cases where users are stuck on a given version of Connect and can't upgrade to 2.6 (or whichever version this KIP lands on) any time soon. 3. The "error.rest.response.message.detail.enabled" property is a bit of a mouthful; it'd be great if we could come up with something more succinct. What do you think about something like "rest.response.stack.traces"? 4. The KIP is targeted at stack traces for uncaught exceptions, but it's also possible that stack traces get exposed in the REST API when querying the status of a connector or one of its tasks. Was this intentional? If so, it'd be great to call out why that kind of filtering is not required in the "Rejected Alternatives" section, and if not, it's probably not too late to consider modifying the KIP to cover those cases as well. 5. The KIP mentions creating a new, separate exception mapper class. This seems like more of an implementation detail and something that can be decided on during code review; unless it's critical to the functionality that the KIP aims to accomplish, I'd suggest leaving that part out since it shouldn't affect the impact on users of the Connect framework. Thanks for the KIP, looking forward to seeing this happen! Cheers, Chris On Thu, Apr 2, 2020 at 11:01 AM Connor Penhale wrote: > Hello All! > > I’ve created the following KIP: > https://cwiki.apache.org/confluence/display/KAFKA/KIP-587:+Suppress+detailed+responses+for+handled+exceptions+in+security-sensitive+environments > > The PR that originated this discussion, is here: > https://github.com/apache/kafka/pull/8355 It is based on 2.0, but I > would be working on Kafka Connect in 2.6 to get this behavior changed to > the community’s preference. > > Looking forward to working with everyone! > > Thanks! > Connor > --- > Connor Penhale | Enterprise Architect, OpenLogic (https://openlogic.com/) > Perforce (https://www.perforce.com/) > Support: +1 866.399.6736 > > > > > This e-mail may contain information that is privileged or confidential. If > you are not the intended recipient, please delete the e-mail and any > attachments and notify us immediately. > > CAUTION: This email originated from outside of the organization. Do not click on links or open
Re: [DISCUSS] KIP-587 Suppress detailed responses for handled exceptions in security-sensitive environments
Hi Connor, Great stuff! I generally like being able to see the stack trace of an exception directly via the REST API but can definitely understand the security concerns here. I've got a few questions/remarks about the KIP and would be interested in your thoughts: 1. The KIP mentions a SUPRESSED_EXCEPTION_MESSAGE, but doesn't actually outline what this message would actually be. It'd be great to see the actual message in the KIP since people may have thoughts on what it should be and want to comment on it as part of this discussion. 2. In the "Rejected Alternatives" section, an Nginx proxy is mentioned as one possible way to filter out stack traces from the REST API. It seems like a Connect REST extension ( https://kafka.apache.org/24/javadoc/index.html?org/apache/kafka/connect/rest/ConnectRestExtension.html) would be a better alternative than an Nginx proxy; had you considered utilizing one? I still think this KIP is worthwhile and a REST extension shouldn't be necessary in order to lock down the REST API this way, but it might be worth calling out as an alternative and perhaps even a workaround in cases where users are stuck on a given version of Connect and can't upgrade to 2.6 (or whichever version this KIP lands on) any time soon. 3. The "error.rest.response.message.detail.enabled" property is a bit of a mouthful; it'd be great if we could come up with something more succinct. What do you think about something like "rest.response.stack.traces"? 4. The KIP is targeted at stack traces for uncaught exceptions, but it's also possible that stack traces get exposed in the REST API when querying the status of a connector or one of its tasks. Was this intentional? If so, it'd be great to call out why that kind of filtering is not required in the "Rejected Alternatives" section, and if not, it's probably not too late to consider modifying the KIP to cover those cases as well. 5. The KIP mentions creating a new, separate exception mapper class. This seems like more of an implementation detail and something that can be decided on during code review; unless it's critical to the functionality that the KIP aims to accomplish, I'd suggest leaving that part out since it shouldn't affect the impact on users of the Connect framework. Thanks for the KIP, looking forward to seeing this happen! Cheers, Chris On Thu, Apr 2, 2020 at 11:01 AM Connor Penhale wrote: > Hello All! > > I’ve created the following KIP: > https://cwiki.apache.org/confluence/display/KAFKA/KIP-587:+Suppress+detailed+responses+for+handled+exceptions+in+security-sensitive+environments > > The PR that originated this discussion, is here: > https://github.com/apache/kafka/pull/8355 It is based on 2.0, but I > would be working on Kafka Connect in 2.6 to get this behavior changed to > the community’s preference. > > Looking forward to working with everyone! > > Thanks! > Connor > --- > Connor Penhale | Enterprise Architect, OpenLogic (https://openlogic.com/) > Perforce (https://www.perforce.com/) > Support: +1 866.399.6736 > > > > > This e-mail may contain information that is privileged or confidential. If > you are not the intended recipient, please delete the e-mail and any > attachments and notify us immediately. > >
