Updates rotateLocked so that we hold the activeStderrWriteForTest write
lock around the dup2Stderr call, rather than acquiring it only after
dup2 was already compelete. This ensures no stderrWriteForTest calls
can race with the dup2 syscall. The now unused waitIdleStderrForTest has
been removed.
On macOS, dup2 and write on the same file descriptor are not atomic with
respect to each other, when rotateLocked called dup2Stderr to redirect
the stderr fd to a new file, concurrent goroutines calling
stderrWriteForTest could observe the fd in a transiently invalid state,
resulting in the bad file descripter.
Fixestailscale/corp#36953
Signed-off-by: James Scott <jim@tailscale.com>
In the event of multiple Filch intances being backed by the same file,
it is possible that concurrent rotateLocked calls occur.
One operation might clear the file,
resulting in another skipping the call to resetReadBuffer,
resulting in a later panic because the read index is invalid.
To at least avoid the panic, always call resetReadBuffer.
Note that the behavior of Filch is undefined when using the same file.
While this avoids the panic, we may still experience data corruption or less.
Fixes#18552
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
This file was never truly necessary and has never actually been used in
the history of Tailscale's open source releases.
A Brief History of AUTHORS files
---
The AUTHORS file was a pattern developed at Google, originally for
Chromium, then adopted by Go and a bunch of other projects. The problem
was that Chromium originally had a copyright line only recognizing
Google as the copyright holder. Because Google (and most open source
projects) do not require copyright assignemnt for contributions, each
contributor maintains their copyright. Some large corporate contributors
then tried to add their own name to the copyright line in the LICENSE
file or in file headers. This quickly becomes unwieldy, and puts a
tremendous burden on anyone building on top of Chromium, since the
license requires that they keep all copyright lines intact.
The compromise was to create an AUTHORS file that would list all of the
copyright holders. The LICENSE file and source file headers would then
include that list by reference, listing the copyright holder as "The
Chromium Authors".
This also become cumbersome to simply keep the file up to date with a
high rate of new contributors. Plus it's not always obvious who the
copyright holder is. Sometimes it is the individual making the
contribution, but many times it may be their employer. There is no way
for the proejct maintainer to know.
Eventually, Google changed their policy to no longer recommend trying to
keep the AUTHORS file up to date proactively, and instead to only add to
it when requested: https://opensource.google/docs/releasing/authors.
They are also clear that:
> Adding contributors to the AUTHORS file is entirely within the
> project's discretion and has no implications for copyright ownership.
It was primarily added to appease a small number of large contributors
that insisted that they be recognized as copyright holders (which was
entirely their right to do). But it's not truly necessary, and not even
the most accurate way of identifying contributors and/or copyright
holders.
In practice, we've never added anyone to our AUTHORS file. It only lists
Tailscale, so it's not really serving any purpose. It also causes
confusion because Tailscalars put the "Tailscale Inc & AUTHORS" header
in other open source repos which don't actually have an AUTHORS file, so
it's ambiguous what that means.
Instead, we just acknowledge that the contributors to Tailscale (whoever
they are) are copyright holders for their individual contributions. We
also have the benefit of using the DCO (developercertificate.org) which
provides some additional certification of their right to make the
contribution.
The source file changes were purely mechanical with:
git ls-files | xargs sed -i -e 's/\(Tailscale Inc &\) AUTHORS/\1 contributors/g'
Updates #cleanup
Change-Id: Ia101a4a3005adb9118051b3416f5a64a4a45987d
Signed-off-by: Will Norris <will@tailscale.com>
Add metrics about logtail uploading and underlying buffer.
Add metrics to the in-memory buffer implementation.
Updates tailscale/corp#21363
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
The filch implementation is fairly broken:
* When Filch.cur exceeds MaxFileSize, it calls moveContents
to copy the entirety of cur into alt (while holding the write lock).
By nature, this is the movement of a lot of data in a hot path,
meaning that all log calls will be globally blocked!
It also means that log uploads will be blocked during the move.
* The implementation of moveContents is buggy in that
it copies data from cur into the start of alt,
but fails to truncate alt to the number of bytes copied.
Consequently, there are unrelated lines near the end,
leading to out-of-order lines when being read back.
* Data filched via stderr do not directly respect MaxFileSize,
which is only checked every 100 Filch.Write calls.
This means that it is possible that the file grows far beyond
the specified max file size before moveContents is called.
* If both log files have data when New is called,
it also copies the entirety of cur into alt.
This can block the startup of a process copying lots of data
before the process can do any useful work.
* TryReadLine is implemented using bufio.Scanner.
Unfortunately, it will choke on any lines longer than
bufio.MaxScanTokenSize, rather than gracefully skip over them.
The re-implementation avoids a lot of these problems
by fundamentally eliminating the need for moveContent.
We enforce MaxFileSize by simply rotating the log files
whenever the current file exceeds MaxFileSize/2.
This is a constant-time operation regardless of file size.
To more gracefully handle lines longer than bufio.MaxScanTokenSize,
we skip over these lines (without growing the read buffer)
and report an error. This allows subsequent lines to be read.
In order to improve debugging, we add a lot of metrics.
Note that the the mechanism of dup2 with stderr
is inherently racy with a the two file approach.
The order of operations during a rotation is carefully chosen
to reduce the race window to be as short as possible.
Thus, this is slightly less racy than before.
Updates tailscale/corp#21363
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
This updates all source files to use a new standard header for copyright
and license declaration. Notably, copyright no longer includes a date,
and we now use the standard SPDX-License-Identifier header.
This commit was done almost entirely mechanically with perl, and then
some minimal manual fixes.
Updates #6865
Signed-off-by: Will Norris <will@tailscale.com>
Scanning log lines is a frequent source of allocations.
Pre-allocate a re-usable buffer.
This still doesn't help when there are giant log lines.
Those will still be problematic from an iOS memory perspective.
For more on that, see https://github.com/tailscale/corp/issues/2423.
(For those who cannot follow that link, it is a discussion
of particular problematic types of log lines for
particular categories of customers. The "categories of customers"
part is the reason that it is a private issue.)
There is also a latent bug here. If we ever encounter
a log line longer than bufio.MaxScanTokenSize,
then bufio.Scan will return an error,
and we'll truncate the file and discard the rest of the log.
That's not good, but bufio.MaxScanTokenSize is really big,
so it probably doesn't matter much in practice now.
Unfortunately, it does prevent us from easily capping the potential
memory usage here, on pain of losing log entries.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
James Scott
Joe Tsai
Will Norris
Maisem Ali
Josh Bleecher Snyder
David Crawshaw
Josh Bleecher Snyder
Brad Fitzpatrick
David Anderson
Earl Lee