Skip to content

Commit

Permalink
updated zfs_info script to handle single device vdevs more reliably
Browse files Browse the repository at this point in the history
  • Loading branch information
@markdhooper
markdhooper committed Nov 29, 2022
1 parent dcd0e84 commit b90804c
Show file tree
Hide file tree
Showing 6 changed files with 194 additions and 83 deletions.
252 changes: 176 additions & 76 deletions 45drives-disks/public/scripts/zfs_info
View file
Original file line number Diff line number Diff line change
Expand Up @@ -79,6 +79,27 @@ def get_zpool_list():

return zpools

def zpool_status_flags(pool_name):
try:
zpool_status_result = subprocess.Popen(
["zpool", "status", pool_name, "-P"], stdout=subprocess.PIPE, stderr=subprocess.DEVNULL, universal_newlines=True).stdout.read()
except:
print(f"failed to run 'zpool status {pool_name}'")
exit(1)
groupings = [(match.group(1),match.group(0)) for match in re.finditer(r"^\t{1}(\S+).*$\n(?:^\t{1} +.*$\n)+|^\t{1}(\S+).*$\n(?:^\t{1} +.*$\n)+",zpool_status_result,flags=re.MULTILINE)]
groupings.append( [("state",match.group(1)) for match in re.finditer(r"^.*state\:\s+(\S+)",zpool_status_result,flags=re.MULTILINE)][0])
return dict(groupings)

def zpool_iostat_flags(pool_name):
try:
zpool_status_result = subprocess.Popen(
["zpool", "iostat","-vP", pool_name], stdout=subprocess.PIPE, stderr=subprocess.DEVNULL, universal_newlines=True).stdout.read()
except:
print(f"failed to run 'zpool iostat -v {pool_name}'")
exit(1)
groupings = [(match.group(1),match.group(0)) for match in re.finditer(r"^(\S+).*$\n(?:^ +.*$\n)+|^(\S+).*$\n(?:^ +.*$\n)+",zpool_status_result,flags=re.MULTILINE)]
return dict(groupings)

def zpool_status(pool_name):
try:
zpool_status_result = subprocess.Popen(
Expand All @@ -104,96 +125,86 @@ def zpool_iostat(pool_name):
groupings = [(match.group(1),match.group(0)) for match in re.finditer(r"^(\S+).*$\n(?:^ +.*$\n)+|^(\S+).*$\n(?:^ +.*$\n)+",zpool_status_result,flags=re.MULTILINE)]
return dict(groupings)

def zpool_status_parse(zp_status_obj,key):
def zpool_status_parse(zp_status_obj,key,pool_name):
if key not in zp_status_obj.keys():
return [], []
vdevs = [
{
"tag":key,
"name": match.group(1),
"state": match.group(2),
"read_errors": match.group(3),
"write_errors": match.group(4),
"checksum_errors": match.group(5)
} for match in re.finditer(r"^\t (\S+-\d+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_obj[key],flags=re.MULTILINE)
]
disks = [
{
"tag":key,
"name":match.group(1),
"state": match.group(2),
"read_errors": match.group(3),
"write_errors": match.group(4),
"checksum_errors": match.group(5)
} for match in re.finditer(r"^\t (\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_obj[key],flags=re.MULTILINE)
]

exception_match = r"^(\d+-\d+)(?:-part[0-9])"
for disk in disks:
match = re.match(exception_match,disk["name"])
if match:
disk["name"] = match.group(1)
#perform zpool status <pool_name> -P command and return the separated object.
zp_status_obj_path = zpool_status_flags(pool_name)

zp_status_default = zp_status_obj[key].splitlines()
zp_status_path = zp_status_obj_path[key].splitlines()

vdevs = []
disks = []
counts = []
disk_count = 0
initial_disk = True
for line in zp_status_obj[key].splitlines():
regex_vdev = re.search("^\t (\S+-\d+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",line)
regex_disk = re.search("^\t (\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",line)
if regex_vdev != None and not initial_disk:
counts.append(disk_count)
disk_count = 0
if regex_disk != None:
for i in range(0,len(zp_status_default)):
re_vdev_default = re.search("^\t (\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_default[i])
re_vdev_path = re.search("^\t (/dev/\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_path[i])
re_disk_path = re.search(f"^\t (/dev/\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_path[i])
re_disk_default = re.search(f"^\t (\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_default[i])
if re_vdev_default != None:
#we have a vdev
vdevs.append(
{
"tag":key,
"name": re_vdev_default.group(1),
"state": re_vdev_default.group(2),
"read_errors": re_vdev_default.group(3),
"write_errors": re_vdev_default.group(4),
"checksum_errors": re_vdev_default.group(5)
}
)
if re_vdev_path != None:
#vdev is also a disk. store it in the array of disks
disks.append(
{
"tag":key,
"name":re_vdev_default.group(1),
"state": re_vdev_default.group(2),
"read_errors": re_vdev_default.group(3),
"write_errors": re_vdev_default.group(4),
"checksum_errors": re_vdev_default.group(5)
}
)
if not initial_disk:
#store the previous disk count, as we need to track the previous vdev's disks
counts.append(disk_count)
disk_count = 1
else:
# this is the first disk encountered, push disk count
disk_count = disk_count +1
initial_disk = False

elif not initial_disk:
#This vdev is not a disk, and we already have encountered a disk before. append the disk count then reset it.
counts.append(disk_count)
disk_count = 0
if re_disk_path != None and re_disk_default != None:
#we've encountered our first regular disk. store the disk, and increment the disk count.
initial_disk = False
disks.append(
{
"tag":key,
"name":re_disk_default.group(1),
"state": re_disk_default.group(2),
"read_errors": re_disk_default.group(3),
"write_errors": re_disk_default.group(4),
"checksum_errors": re_disk_default.group(5)
}
)
disk_count = disk_count + 1
counts.append(disk_count)

return vdevs, disks, counts

def zpool_iostat_parse(zp_status_obj,key):
if key not in zp_status_obj.keys():
return [], []
vdevs = [
{
"tag":key,
"raid_level":match.group(1),
"alloc": match.group(2),
"free": match.group(3),
"read_ops": match.group(4),
"write_ops": match.group(5),
"read_bw": match.group(6),
"write_bw": match.group(7)
} for match in re.finditer(r"^ (\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_obj[key],flags=re.MULTILINE)
]
disks = [
{
"tag":key,
"name":match.group(1),
"alloc": match.group(2),
"free": match.group(3),
"read_ops": match.group(4),
"write_ops": match.group(5),
"read_bw": match.group(6),
"write_bw": match.group(7)
} for match in re.finditer(r"^ (\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_status_obj[key],flags=re.MULTILINE)
]

#implement the fix for legacy true-nas customers (they had part-2 as part of the disk name):
exception_match = r"^(\d+-\d+)(?:-part[0-9])"
for disk in disks:
match = re.match(exception_match,disk["name"])
if match:
disk["name"] = match.group(1)

counts = []
disk_count = 0
for line in zp_status_obj[key].splitlines():
regex_vdev = re.search("^ (\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",line)
regex_disk = re.search("^ (\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",line)
if regex_vdev != None and disk_count > 0:
counts.append(disk_count)
disk_count = 0
if regex_disk != None:
disk_count = disk_count + 1
#append the final disk count to the array and return both arrays and counts.
counts.append(disk_count)
return vdevs, disks, counts

Expand Down Expand Up @@ -242,6 +253,95 @@ def verify_zfs_device_format(zp_status_obj,pool_name):
#sys.exit(1)
return alert

def zpool_iostat_parse(zp_iostat_obj,key,pool_name):
if key not in zp_iostat_obj.keys():
return [], []

#perform zpool iostat -vP <pool_name> command and return the separated object.
zp_iostat_obj_path = zpool_iostat_flags(pool_name)

zp_iostat_default = zp_iostat_obj[key].splitlines()
zp_iostat_path = zp_iostat_obj_path[key].splitlines()

vdevs = []
disks = []
counts = []
disk_count = 0
initial_disk = True
for i in range(0,len(zp_iostat_default)):
re_vdev_default = re.search(r"^ (\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_iostat_default[i])
re_vdev_path = re.search(r"^ (/dev/\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_iostat_path[i])
re_disk_path = re.search(r"^ (/dev/\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_iostat_path[i])
re_disk_default = re.search(r"^ (\S+)(?:-part[0-9])?\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+).*",zp_iostat_default[i])
if re_vdev_default != None:
#we have a vdev
vdevs.append(
{
"tag":key,
"raid_level":re_vdev_default.group(1),
"alloc": re_vdev_default.group(2),
"free": re_vdev_default.group(3),
"read_ops": re_vdev_default.group(4),
"write_ops": re_vdev_default.group(5),
"read_bw": re_vdev_default.group(6),
"write_bw": re_vdev_default.group(7)
}
)
if re_vdev_path != None:
#vdev is also a disk. store it in the array of disks, also update the raid_level of the vdev to "Disk".
disks.append(
{
"tag":key,
"name":re_vdev_default.group(1),
"alloc": re_vdev_default.group(2),
"free": re_vdev_default.group(3),
"read_ops": re_vdev_default.group(4),
"write_ops": re_vdev_default.group(5),
"read_bw": re_vdev_default.group(6),
"write_bw": re_vdev_default.group(7)
}
)
vdevs[-1]["raid_level"] = "Disk"
if not initial_disk:
#store the previous disk count, as we need to track the previous vdev's disks
counts.append(disk_count)
#a vdev that is also a disk will only have one disk.
disk_count = 1
else:
# this is the first disk encountered, increment the disk count and clear the flag.
disk_count = disk_count + 1
initial_disk=False
elif not initial_disk:
#This vdev is not a disk, and we already have encountered a disk before. append the disk count then reset it.
counts.append(disk_count)
disk_count = 0
if re_disk_path != None and re_disk_default != None:
#we've encountered our first regular disk. store the disk, and increment the disk count.
initial_disk = False
disks.append(
{
"tag":key,
"name":re_disk_default.group(1),
"alloc": re_disk_default.group(2),
"free": re_disk_default.group(3),
"read_ops": re_disk_default.group(4),
"write_ops": re_disk_default.group(5),
"read_bw": re_disk_default.group(6),
"write_bw": re_disk_default.group(7)
}
)
disk_count = disk_count + 1

#implement the fix for legacy nfs customers (they had part-2 as part of the disk name):
exception_match = r"^(\d+-\d+)(?:-part[0-9])"
for disk in disks:
match = re.match(exception_match,disk["name"])
if match:
disk["name"] = match.group(1)

#append the final disk count to the array and return both arrays and counts.
counts.append(disk_count)
return vdevs, disks, counts

def get_zpool_status():
json_zfs["warnings"] = []
Expand All @@ -258,8 +358,8 @@ def get_zpool_status():
# parse the output of both commands by top level entry
if key in iostat_output.keys():
# get all parsed output as arrays of objects from each command.
status_vdevs, status_disks, status_disk_counts = zpool_status_parse(status_output,key)
iostat_vdevs, iostat_disks, iostat_disk_counts = zpool_iostat_parse(iostat_output,key)
status_vdevs, status_disks, status_disk_counts = zpool_status_parse(status_output,key,pool["name"])
iostat_vdevs, iostat_disks, iostat_disk_counts = zpool_iostat_parse(iostat_output,key,pool["name"])
if not status_disks or not iostat_disks or not status_disk_counts or not iostat_disk_counts:
print("/usr/share/cockpit/45drives-disks/scripts/zfs_info failed to interpret the following zfs information:")
print("zpool status {pn}:".format(pn=pool["name"]))
Expand Down
5 changes: 3 additions & 2 deletions CHANGELOG.md
View file
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
## cockpit-45drives-hardware 2.2.1-6
## cockpit-45drives-hardware 2.2.2-1

* fixed drive placement error in 45drives-disks for XL60-H32 units not displaying disk 4-10 properly
* updated zfs_info script to fix bug when single disk vdevs are encountered
* single device vdevs are now reported as both a disk and vdev by zfs_info script
2 changes: 1 addition & 1 deletion Makefile
View file
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@
PLUGIN_SRCS=

# For installing to a remote machine for testing with `make install-remote`
REMOTE_TEST_HOST=192.168.35.36
REMOTE_TEST_HOST=192.168.13.33
REMOTE_TEST_USER=root

# Restarts cockpit after install
Expand Down
8 changes: 4 additions & 4 deletions manifest.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,8 +3,8 @@
"name": "cockpit-45drives-hardware",
"title": "cockpit-45drives-hardware",
"prerelease": false,
"version": "2.2.1",
"buildVersion": "6",
"version": "2.2.2",
"buildVersion": "1",
"author": "Mark Hooper <[email protected]>",
"url": "https://github.com/45Drives/cockpit-hardware",
"category": "utils",
Expand Down Expand Up @@ -58,8 +58,8 @@
],
"changelog": {
"urgency": "medium",
"version": "2.2.1",
"buildVersion": "6",
"version": "2.2.2",
"buildVersion": "1",
"ignore": [],
"date": null,
"packager": "Mark Hooper <[email protected]>",
Expand Down
3 changes: 3 additions & 0 deletions packaging/el8/main.spec
View file
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,9 @@ make DESTDIR=%{buildroot} install
/usr/lib/udev/rules.d/68-cockpit-45drives-disks.rules

%changelog
* Tue Nov 29 2022 Mark Hooper <[email protected]> 2.2.2-1
- updated zfs_info script to fix bug when single disk vdevs are encountered
- single device vdevs are now reported as both a disk and vdev by zfs_info script
* Mon Oct 31 2022 Mark Hooper <[email protected]> 2.2.1-6
- fixed drive placement error in 45drives-disks for XL60-H32 units not displaying
disk 4-10 properly
Expand Down
7 changes: 7 additions & 0 deletions packaging/focal/changelog
View file
Original file line number Diff line number Diff line change
@@ -1,3 +1,10 @@
cockpit-45drives-hardware (2.2.2-1focal) focal; urgency=medium

* updated zfs_info script to fix bug when single disk vdevs are encountered
* single device vdevs are now reported as both a disk and vdev by zfs_info script

-- Mark Hooper <[email protected]> Tue, 29 Nov 2022 12:30:50 -0400

cockpit-45drives-hardware (2.2.1-6focal) focal; urgency=medium

* fixed drive placement error in 45drives-disks for XL60-H32 units not displaying
Expand Down

0 comments on commit b90804c

Please sign in to comment.