Commit Graph

37 Commits (fc73586c971225037aa09b5462031b9694278c74)

Author SHA1 Message Date
dt-rush 5d3e2ce2ef properly strip path.rootfs from mountpoint labels (#1421)
Change-type: patch
Connects-to: #1418
Signed-off-by: dt-rush <nickp@balena.io>
2019-07-19 16:51:17 +02:00
Leonid Evdokimov 22a7dbae08 Ignore iso9600 filesystem on Linux (#1355)
The filesystem is read-only and is often used for a virtual FS
with a configuration file for a virtual machine.

Signed-off-by: Leonid Evdokimov <leon@darkk.net.ru>
2019-06-18 17:47:05 +01:00
Paul Gier d0a66c4c40 use sys/unix package instead of syscall (#1340)
According to the golang docs, the syscall package is deprecated.
https://golang.org/pkg/syscall
This updates collectors to use the x/sys/unix package instead.
Also updates the vendored x/sys/unix module to latest.

Signed-off-by: Paul Gier <pgier@redhat.com>
2019-05-10 13:04:06 -05:00
Ben Kochie 4abc6fba7d
Add fallback for missing /proc/1/mounts (#1172)
* Add fallback for missing /proc/1/mounts

On some systems, `/proc/1/mounts` is hidden from non-root users due to
the `hidepid` procfs feature. Attempt to fallback to `/proc/mounts` if
`/proc/1/mounts` is not found.

Signed-off-by: Ben Kochie <superq@gmail.com>

* Add tests.

Signed-off-by: Ben Kochie <superq@gmail.com>

* Add CHANGELOG entry.

Signed-off-by: Ben Kochie <superq@gmail.com>
2018-11-30 14:01:55 +01:00
Jerome Froelich 0cb0c4d911 Remove unused variable readOnly from filesystem_linux.go. (#1173)
The pull request #1002 changed the logic used on Linux servers to determine if a filesystem is
read-only. As a result of this change, the variable `readOnly` is now unused and can be removed.

Signed-off-by: Jerome Froelich <jeromefroelich@hotmail.com>
2018-11-30 14:01:39 +01:00
Matt Layher c0a55e3f80 collector: add bounds check and test for filesystem collector (#1133)
Signed-off-by: Matt Layher <mdlayher@gmail.com>
2018-10-30 22:12:42 +01:00
Ben Kochie 0880d460d7
Ignore additional virtual filesystems (#1104)
Add more virtual filesystems to the default ignore list
* bpf
* cgroup2
* selinuxfs
* squashfs

Signed-off-by: Ben Kochie <superq@gmail.com>
2018-10-12 11:24:32 +02:00
Yecheng Fu 0f9842f20a [continue 912] strip rootfs prefix for run in docker (#1058)
* strip rootfs prefix for run in docker
* Use `/` as default value of path.rootfs, and parse mounts from `/proc/1/mounts`.
* No need to mount `/proc` and `/sys` because we share host's PID
namespace, which allows processes within the container to see all of the
processes on the system.

Closes: #66

Signed-off-by: Ivan Mikheykin <ivan.mikheykin@flant.com>
Signed-off-by: Yecheng Fu <cofyc.jackson@gmail.com>
2018-10-04 14:11:21 +02:00
Daniele Sluijters d999dacdc6 filesystem: Ignore netns/nsfs mounts (#1047)
When starting Docker containers a whole bunch of netns (network
namespace) mounts are created that the node exporter can't make any
sense of (and can't read either).

This ignores all nsfs filesystems.

Fixes #875

Signed-off-by: Daniele Sluijters <daenney@users.noreply.github.com>
2018-09-26 10:45:51 +02:00
Johannes Wienke 5c780d132c Exclude only subdirectories of /var/lib/docker (#1003)
It is quite common to put /var/lib/docker itself on a separate partition
and that should be monitored as well.

Signed-off-by: Johannes Wienke <languitar@semipol.de>
2018-07-23 15:43:42 +02:00
Brandon Gilmore 76bbd8dd18 Use /proc/mounts instead of statfs(2) for ro state (#1002)
While the statfs(2) approach is reliable for normally mounted filesystems, the
flags returned can be inconsistent when filesystem has been remounted read-only
after encountering an error. The returned flags do accurately represent the
internal state of the filesystem, but they do not reflect whether the VFS layer
will accept writes. Instead, it makes sense to parse the current VFS mount
state from the options field in /proc/mounts since it takes precedence.

Signed-off-by: Brandon Gilmore <bgilmore@valvesoftware.com>
2018-07-16 15:56:27 +02:00
mknapphrt 09b4305090 Changed the way that stuck mounts are handled. If a mount fails to return, it will stop being queried until it returns. (#997)
Fixed spelling mistakes.

Update transport_generic.go

Changed to a mutex approach instead of channels and added a timeout before declaring a mount stuck.

Removed unnecessary lock channel and clarified some var names.

Fixed style nits.

Signed-off-by: Mark Knapp <mknapp@hudson-trading.com>
2018-07-14 11:10:28 +02:00
Brad Beam e3cf1d5187 Adding support for evaluating octal characters in mountpoint (#954)
Signed-off-by: Brad Beam <brad.beam@b-rad.info>
2018-06-06 16:49:19 +02:00
Kasinath Kottukkal f6965e1812 Add overlay to defIgnoredFSTypes (#824)
* Add overlay to defIgnoredFSTypes

To avoid statfs() errors if node_exporter is running as non privileged user.

* Updated defIngoredFSTypes values in sorted order
2018-02-16 09:47:50 +01:00
Brian Brazil 1072f2868d Fix log level regression in #533 2018-02-07 15:16:20 +00:00
Brian Brazil 7e41a2b279 Ignore /var/lib/docker by default. (#814)
The node exporter runs unprivileged, so it cannot statfs any filesystems
under this directory causing log spam.  In addition there tends to be
high churn in the filesystems here (as it's basically application
monitoring) which can cause high cardinaltiy and in one case caused
Prometheus's index symbol table to get very large.
Accordingly this should be ignored to reduce log spam and avoid
performance issues. The filesystems themselves can in principle be
monitored via container oriented exporters, and the underlying
filesystems will still be monitored.
2018-02-06 17:10:59 +01:00
Ben Kochie 7f6c59e198
Ignore more virtual filesystems (#775)
Add additional Linux virtual filesystem types to the default list.
2018-01-03 17:22:02 +01:00
Netmonk 2aa8d0eb0c [FIX] Exclude Linux proc from filesystem type regexp (#774)
* [FIX] Issue 63, error on excluding proc filesystem on linux, improving regexp

* [FIX] Reordering filter order
2018-01-03 11:40:32 +01:00
Tobias Schmidt d290ea94b8 Fix export of stale device error metrics for unmounted filesystems
Instead of maintaining a counter metric for device errors in memory,
this change exports a gauge and uses const metrics to avoid leaking
metrics for unmounted filesystems.
2017-03-22 21:48:18 -03:00
Tobias Schmidt 922e74d58f Remove unnecessarily named return variables
Named return variables should only be used to describe the returned type
further, e.g. `err error` doesn't add any new information and is just
stutter.
2017-02-28 16:04:25 -04:00
Tobias Schmidt c703435790 Fix all open go lint and vet issues 2017-02-28 13:05:38 -04:00
Johannes 'fish' Ziemke 4c9131b7d8 Make sure we only return one metric per mounted fs 2017-01-04 16:45:25 +01:00
Johannes 'fish' Ziemke d506b2266c Merge pull request #374 from prometheus/fish-add-filesystem-errors
Add node_filesystem_device_errors_total metric
2016-12-26 11:51:14 +01:00
Bjørn Forsman 64e637cbcc Ignore autofs filesystems on linux
node_exporter currently triggers autofs to mount the underlying
filesystem on every scrape. This is undesirable. Better ignore autofs.

The underlying filesystem that autofs mounts will be monitored though,
when the (real) filesystem is mounted.
2016-12-25 15:13:45 +01:00
Johannes 'fish' Ziemke deebf0aa49 Add node_filesystem_device_errors_total metric
This metric is the total number of errors occurred when getting stats
for the given device.
2016-12-19 11:48:32 +01:00
Sharif Nassar bf88b6ee0f Fix typos, and initialisms from comments in #217 2016-05-13 13:40:14 -07:00
Dominik Schulz 3fe5e7f8fa Add ignored fs types 2016-05-13 13:40:14 -07:00
Tobias Schmidt 041de0e30c Fix go format style in filesystem collector 2016-01-21 16:20:32 -05:00
Will Rouesnel 05539ee156 Add filesystem read-only metric node_filesystem_readonly
This is a boolean metric which is set to 1 when the filesystem is flagged as
read-only.
2015-11-12 14:11:07 +11:00
Nick Owens ebf3ef32f1 collector: sync bsd filesystem collector labels with linux 2015-11-02 19:16:17 -08:00
Nick Owens eb79937340 switch to github.com/prometheus/common/log for logging 2015-10-30 13:20:06 -07:00
Kevin Lyda 36291d456e Make filesystem collector more robust.
Current behaviour throws away all stats on any Statfs error. In practice
this is not useful. This turns such errors into debug log messages -
though silently ignoring them seems even more valid to me.
2015-10-18 19:47:45 +01:00
Matthias Rampke 2d0d72b97d Add license headers to all code files. 2015-09-26 17:44:39 +02:00
Matthias Rampke 20b551ab2b Add flag to change the location of the procfs.
Remove all hardcoded references to `/proc`. For all collectors that do
not use `github.com/prometheus/procfs` yet, provide a wrapper to
generate the full paths.

Reformulate help strings, errors and comments to remove absolute
references to `/proc`.

This is a breaking change: the `-collector.ipvs.procfs` flag is removed
in favor of the general flag. Since it only affected that collector it
was only useful for development, so this should not cause many issues.
2015-09-26 15:32:10 +02:00
Martín Ferrari dace6dc48b Fix issues from code review, and split common code into filesystem_common.go. 2015-09-16 16:34:34 +03:00
Martín Ferrari 8c2316e8a9 Revamp the filesystem collector to use throw-away ConstMetrics. 2015-09-15 12:17:15 +00:00
Siavash Safi 66e0171159 Rename linux collectors for build constraints. 2015-07-14 13:58:45 +04:30