Adding a Debug reader to detect Null Bytes from the io.Reader #9210
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@urso This is briefly what we discussed yesterday about null-bytes over network volume or the issue that we are still investigating at https://discuss.elastic.co/t/filbeat-still-ooms/156211/25 It should be pretty low impact even in production, if we detect too much null-bytes the reader will just disable itself. But in the log at info level we will see that we have matched some bytes + surrounding data. QuestionsShould we log it at info level or should we enforce a special selector? |
ph
force-pushed
the
debug/large-buffer-issue-copy
branch
from
6377408 to
009e940
Compare
|
Error is legit under windows, looking into it. |
urso
reviewed
Nov 22, 2018
ph
force-pushed
the
debug/large-buffer-issue-copy
branch
2 times, most recently
from
65ad07f to
452fa40
Compare
|
@urso I think I have addressed all the issues in the PR, the |
urso
suggested changes
Nov 23, 2018
|
@urso updated with the latest comments. |
|
Oh windows, I presume I am not correctly writing the null bytes on windows. |
urso
approved these changes
Nov 23, 2018
|
LGTM |
When you are reading logs on a network volume like NFS or ZFS it is possible that the underlying filesystem return null bytes instead of returning concrete data, its not currently possible to detect that in all scenario unless events are eventually send to ES and you can inspect them and see \u0000 chars in the messages. This is a small proposal to add a Debug Reader which should only by used for debugging purpose it allow to log if any null bytes are present in the streams of bytes and will log surround values. It accepts an io.Reader as the source of data, a buffer size, a predicate to check the value of a byte and how much detection invokation it should do before disabling the check. Enable it with either of the following selectors: "*" or "detect_null_bytes"
ph
force-pushed
the
debug/large-buffer-issue-copy
branch
from
3449f3f to
253ff2a
Compare
|
I've fixed the windows issue and rebase the content of this PR, when its green I will merge it. |
ph
added a commit
to ph/beats
that referenced
this pull request
Nov 26, 2018
…c#9210) When you are reading logs on a network volume like NFS or ZFS it is possible that the underlying filesystem return null bytes instead of returning concrete data, its not currently possible to detect that in all scenario unless events are eventually send to ES and you can inspect them and see \u0000 chars in the messages. This is a small proposal to add a Debug Reader which should only by used for debugging purpose it allow to log if any null bytes are present in the streams of bytes and will log surround values. It accepts an io.Reader as the source of data, a buffer size, a predicate to check the value of a byte and how much detection invokation it should do before disabling the check. Enable it with either of the following selectors: "*" or "detect_null_bytes" (cherry picked from commit 96c924a)
ph
added
v6.6.0
and removed
needs_backport
ph
added a commit
that referenced
this pull request
Nov 26, 2018
…from the io.Reader (#9236) * Adding a Debug reader to detect Null Bytes from the io.Reader (#9210) When you are reading logs on a network volume like NFS or ZFS it is possible that the underlying filesystem return null bytes instead of returning concrete data, its not currently possible to detect that in all scenario unless events are eventually send to ES and you can inspect them and see \u0000 chars in the messages. This is a small proposal to add a Debug Reader which should only by used for debugging purpose it allow to log if any null bytes are present in the streams of bytes and will log surround values. It accepts an io.Reader as the source of data, a buffer size, a predicate to check the value of a byte and how much detection invokation it should do before disabling the check. Enable it with either of the following selectors: "*" or "detect_null_bytes" (cherry picked from commit 96c924a)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
When you are reading logs on a network volume like NFS or ZFS it is
possible that the underlying filesystem return null bytes instead of
returning concrete data, its not currently possible to detect that in all
scenario unless events are eventually send to ES and you can inspect
them and see \u0000 chars in the messages.
This is a small proposal to add a Debug Reader which should only by used
for debugging purpose it allow to log if any null bytes are present in
the streams of bytes and will log surround values.
It accepts an io.Reader as the source of data, a buffer size, a
predicate to check the value of a byte and how much detection invokation it should do
before disabling the check.
Enable it with either of the following selectors: "*" or "detect_null_bytes"