nbdkit-error-filter - inject errors for testing clients
nbdkit --filter=error PLUGIN [error=EPERM|EIO|ENOMEM|EINVAL|ENOSPC|ESHUTDOWN] [error-rate=10%|0.1] [error-file=/tmp/inject] [error-pread=...] [error-pread-rate=...] [error-pread-file=...] [error-pwrite=...] [error-pwrite-rate=...] [error-pwrite-file=...] [error-trim=...] [error-trim-rate=...] [error-trim-file=...] [error-zero=...] [error-zero-rate=...] [error-zero-file=...] [error-extents=...] [error-extents-rate=...] [error-extents-file=...] [error-cache=...] [error-cache-rate=...] [error-cache-file=...]
nbdkit-error-filter is an nbdkit filter that injects random errors into replies from the server. This is used for testing that NBD clients can handle errors.
All parameters are optional, but you should usually specify one of the
error-*-rate parameters, otherwise this filter will do nothing.
Inject a low rate of errors randomly into the connection:
nbdkit --filter=error file disk.img error-rate=1%
Reading, trimming, cache and extents (block status) requests will be successful, but all writes and zeroing will return "No space left on device":
nbdkit --filter=error file disk.img \ error=ENOSPC \ error-pwrite-rate=100% \ error-zero-rate=100%
To make all connections fail hard 60 seconds after the server is started, use:
rm -f /tmp/inject nbdkit --filter=error file disk.img \ error-rate=100% \ error-file=/tmp/inject sleep 60; touch /tmp/inject
When a random error is injected, you can select which one from the range of possible NBD errors (the NBD protocol only supports a limited range of error codes).
This parameter is optional and the default is
EIO ("Input/output error").
The rate of injected errors per NBD request. This can be expressed as either a percentage between
100% or as a probability between
0 is used then no errors are ever injected, and if
1 is used then all requests return errors.
This parameter is optional and the default is
0%. Unless you set this, the filter will do nothing.
Errors will only be injected when FILENAME exists. (Note you must also specify the
You can use this for fine-grained control over when to inject errors, for example if you want to trigger an error at an exact moment during a test, arrange for this file to be created at the appropriate time. Or conversely to test error recovery in a client, create the file initially, and then delete it to check the client can recover.
This parameter is optional.
error-file but only apply the settings to NBD pread requests.
error-file but only apply the settings to NBD pwrite requests.
error-file but only apply the settings to NBD trim requests.
error-file but only apply the settings to NBD zero requests.
(nbdkit ≥ 1.12)
error-file but only apply the settings to NBD block status requests to read extents.
(nbdkit ≥ 1.14)
error-file but only apply the settings to NBD cache requests.
If you are looking at the debugging output (using
nbdkit -f -v) references to the name of this filter show up as
"error-inject:", and such lines indicate that the filter is not altering output, for example:
nbdkit: file.9: debug: error-inject: pread count=1024 offset=0 flags=0x0
Conversely, references to the string
"error:" occur when the nbdkit_error(3) API was used, including when this filter injects an error, as in:
nbdkit: file.4: error: injecting ENOSPC error into pwrite
nbdkit --dump-config to find the location of
nbdkit-error-filter first appeared in nbdkit 1.6.
nbdkit(1), nbdkit-file-plugin(1), nbdkit-full-plugin(1), nbdkit-retry-filter(1), nbdkit-retry-request-filter(1), nbdkit-filter(3).
Richard W.M. Jones
Copyright (C) 2018 Red Hat Inc.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
Neither the name of Red Hat nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY RED HAT AND CONTRIBUTORS ''AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL RED HAT OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.