The devstat API expects us to reuse one devinfo for many invocations of
devstat_getstats. In particular, it allocates and resizes memory
referenced by devinfo.
Querying the number of devices separately from the device list itself is
racy. Devices may be added or removed between the two calls; and removed
devices would lead to a segfault.
The memory allocated by calloc was never freed. Since the devinfo struct
never leaves the function, anyway, we might as well just allocate it on
the stack.
It seems solaris prefers "sys/loadavg.h" over "stdlib.h" when
fetching the load average.
For Illumos based OSes it was required to include "sys/time.h" to
ensure that "hrtime_t" was defined.
https://www.illumos.org/issues/6002
It also required setting the ldflags "-fno-stack-protector -lssp" to
avoid undefined symbols when linking with gcc.
/opt/local/go/pkg/tool/solaris_amd64/link: running gcc failed: exit status 1
Undefined first referenced
symbol in file
__stack_chk_fail /tmp/go-link-138622936/000002.o
__stack_chk_guard /tmp/go-link-138622936/000002.o
Instead of doing the whole metric exposition in a platform specific collector
implementation, this creates and updates the metrics in meminfo.go and
expected a platform specific implementation of getMemInfo on
*meminfoCollector.
This removes some error handling, which should be fine. If the calls
fail, we will get the zeroes, which is a safe enough fallback.
Additionally, if the first sysctl (page_size) succeeded it is unlikely
that other ones will fail.
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.
They get printed all the time, as there are some tokens in the /proc
file that we simply don't support. It's better to keep these as
debugging messages, which may come in useful if new tags start to
appear.
Collect metrics from the StorCLI utility on the health of MegaRAID
hardware RAID controllers and write them to stdout so that they can be
used by the textfile collector.
We parse the JSON output that StorCLI provides.
Script must be run as root or with appropriate capabilities for storcli
to access the RAID card.
Designed to run under Python 2.7, using the system Python provided with
many Linux distributions.
The metrics look like this:
mbostock@host:~$ sudo ./storcli.py
megaraid_status_code 0
megaraid_controllers_count 1
megaraid_emergency_hot_spare{controller="0"} 1
megaraid_scheduled_patrol_read{controller="0"} 1
megaraid_virtual_drives{controller="0"} 1
megaraid_drive_groups{controller="0"} 1
megaraid_virtual_drives_optimal{controller="0"} 1
megaraid_degraded{controller="0"} 0
megaraid_battery_backup_healthy{controller="0"} 1
megaraid_ports{controller="0"} 8
megaraid_failed{controller="0"} 0
megaraid_drive_groups_optimal{controller="0"} 1
megaraid_healthy{controller="0"} 1
megaraid_physical_drives{controller="0"} 24
megaraid_controller_info{controller="0", model="AVAGOMegaRAIDSASPCIExpressROMB"} 1
mbostock@host:~$
- Use the right number of printf() arguments. Use %q where it makes sense.
- Use "DRBD" instead of "Drbd", per Go's style guide.
- Add _total suffixes to counter metrics.
- Mention the unit (bytes) in documentation strings once more.
