Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add config options for loki dashboards #2617

Merged
merged 1 commit into from
Sep 18, 2020
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Add config options for loki dashboards
periklis committed Sep 16, 2020
commit e004609408b97e63125c34bf8a908b76a79b19bb
217 changes: 5 additions & 212 deletions production/loki-mixin/dashboards.libsonnet
Original file line number Diff line number Diff line change
@@ -1,212 +1,5 @@
local g = import 'grafana-builder/grafana.libsonnet';
local utils = import 'mixin-utils/utils.libsonnet';

{
grafanaDashboards+: {
'loki-logs.json': import './dashboard-loki-logs.json',
'loki-operational.json': import './dashboard-loki-operational.json',
'loki-writes.json':
g.dashboard('Loki / Writes')
.addMultiTemplate('cluster', 'kube_pod_container_info{image=~".*loki.*"}', 'cluster')
.addMultiTemplate('namespace', 'kube_pod_container_info{image=~".*loki.*"}', 'namespace')
.addRow(
g.row('Frontend (cortex_gw)')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster=~"$cluster", job=~"($namespace)/cortex-gw", route=~"api_prom_push|loki_api_v1_push"}')
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.re('job', '($namespace)/cortex-gw'), utils.selector.re('route', 'api_prom_push|loki_api_v1_push')], extra_selectors=[utils.selector.re('cluster', '$cluster')])
)
)
.addRow(
g.row('Distributor')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster=~"($cluster)", job=~"($namespace)/distributor"}')
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.re('job', '($namespace)/distributor')], extra_selectors=[utils.selector.re('cluster', '$cluster')])
)
)
.addRow(
g.row('Ingester')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster=~"$cluster", job=~"($namespace)/ingester",route="/logproto.Pusher/Push"}')
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.re('job', '($namespace)/ingester'), utils.selector.eq('route', '/logproto.Pusher/Push')], extra_selectors=[utils.selector.re('cluster', '$cluster')])
)
)
.addRow(
g.row('BigTable')
.addPanel(
g.panel('QPS') +
g.qpsPanel('cortex_bigtable_request_duration_seconds_count{cluster=~"$cluster", job=~"($namespace)/ingester", operation="/google.bigtable.v2.Bigtable/MutateRows"}')
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('cortex_bigtable_request_duration_seconds', [utils.selector.re('cluster', '$cluster'), utils.selector.re('job', '($namespace)/ingester')] + [utils.selector.eq('operation', '/google.bigtable.v2.Bigtable/MutateRows')])
)
)
.addRow(
g.row('BoltDB Shipper')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_boltdb_shipper_request_duration_seconds_count{cluster=~"$cluster", job=~"($namespace)/ingester", operation="WRITE"}')
)
.addPanel(
g.panel('Latency') +
g.latencyPanel('loki_boltdb_shipper_request_duration_seconds', '{cluster=~"$cluster", job=~"($namespace)/ingester", operation="WRITE"}')
)
),

local http_routes = 'loki_api_v1_series|api_prom_series|api_prom_query|api_prom_label|api_prom_label_name_values|loki_api_v1_query|loki_api_v1_query_range|loki_api_v1_labels|loki_api_v1_label_name_values',
local grpc_routes = '/logproto.Querier/Query|/logproto.Querier/Label|/logproto.Querier/Series',

'loki-reads.json':
g.dashboard('Loki / Reads')
.addTemplate('cluster', 'kube_pod_container_info{image=~".*loki.*"}', 'cluster')
.addTemplate('namespace', 'kube_pod_container_info{image=~".*loki.*"}', 'namespace')
.addRow(
g.row('Frontend (cortex_gw)')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster="$cluster", job="$namespace/cortex-gw", route=~"%s"}' % http_routes)
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.eq('job', '$namespace/cortex-gw'), utils.selector.re('route', http_routes)], extra_selectors=[utils.selector.eq('cluster', '$cluster')], sum_by=['route'])
)
)
.addRow(
g.row('Frontend (query-frontend)')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster="$cluster", job="$namespace/query-frontend", route=~"%s"}' % http_routes)
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.eq('job', '$namespace/query-frontend'), utils.selector.re('route', http_routes)], extra_selectors=[utils.selector.eq('cluster', '$cluster')], sum_by=['route'])
)
)
.addRow(
g.row('Querier')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster="$cluster", job="$namespace/querier", route=~"%s"}' % http_routes)
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.eq('job', '$namespace/querier'), utils.selector.re('route', http_routes)], extra_selectors=[utils.selector.eq('cluster', '$cluster')], sum_by=['route'])
)
)
.addRow(
g.row('Ingester')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_request_duration_seconds_count{cluster="$cluster", job="$namespace/ingester",route=~"%s"}' % grpc_routes)
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('loki_request_duration_seconds', [utils.selector.eq('job', '$namespace/ingester'), utils.selector.re('route', grpc_routes)], extra_selectors=[utils.selector.eq('cluster', '$cluster')], sum_by=['route'])
)
)
.addRow(
g.row('BigTable')
.addPanel(
g.panel('QPS') +
g.qpsPanel('cortex_bigtable_request_duration_seconds_count{cluster=~"$cluster", job=~"($namespace)/querier", operation="/google.bigtable.v2.Bigtable/ReadRows"}')
)
.addPanel(
g.panel('Latency') +
utils.latencyRecordingRulePanel('cortex_bigtable_request_duration_seconds', [utils.selector.re('cluster', '$cluster'), utils.selector.re('job', '($namespace)/querier')] + [utils.selector.eq('operation', '/google.bigtable.v2.Bigtable/ReadRows')])
)
)
.addRow(
g.row('BoltDB Shipper')
.addPanel(
g.panel('QPS') +
g.qpsPanel('loki_boltdb_shipper_request_duration_seconds_count{cluster=~"$cluster", job=~"($namespace)/querier", operation="QUERY"}')
)
.addPanel(
g.panel('Latency') +
g.latencyPanel('loki_boltdb_shipper_request_duration_seconds', '{cluster=~"$cluster", job=~"($namespace)/querier", operation="QUERY"}')
)
),


'loki-chunks.json':
g.dashboard('Loki / Chunks')
.addTemplate('cluster', 'kube_pod_container_info{image=~".*loki.*"}', 'cluster')
.addTemplate('namespace', 'kube_pod_container_info{image=~".*loki.*"}', 'namespace')
.addRow(
g.row('Active Series / Chunks')
.addPanel(
g.panel('Series') +
g.queryPanel('sum(loki_ingester_memory_chunks{cluster="$cluster", job="$namespace/ingester"})', 'series'),
)
.addPanel(
g.panel('Chunks per series') +
g.queryPanel('sum(loki_ingester_memory_chunks{cluster="$cluster", job="$namespace/ingester"}) / sum(loki_ingester_memory_streams{job="$namespace/ingester"})', 'chunks'),
)
)
.addRow(
g.row('Flush Stats')
.addPanel(
g.panel('Utilization') +
g.latencyPanel('loki_ingester_chunk_utilization', '{cluster="$cluster", job="$namespace/ingester"}', multiplier='1') +
{ yaxes: g.yaxes('percentunit') },
)
.addPanel(
g.panel('Age') +
g.latencyPanel('loki_ingester_chunk_age_seconds', '{cluster="$cluster", job="$namespace/ingester"}'),
),
)
.addRow(
g.row('Flush Stats')
.addPanel(
g.panel('Size') +
g.latencyPanel('loki_ingester_chunk_entries', '{cluster="$cluster", job="$namespace/ingester"}', multiplier='1') +
{ yaxes: g.yaxes('short') },
)
.addPanel(
g.panel('Entries') +
g.queryPanel('sum(rate(cortex_chunk_store_index_entries_per_chunk_sum{cluster="$cluster", job="$namespace/ingester"}[5m])) / sum(rate(cortex_chunk_store_index_entries_per_chunk_count{cluster="$cluster", job="$namespace/ingester"}[5m]))', 'entries'),
),
)
.addRow(
g.row('Flush Stats')
.addPanel(
g.panel('Queue Length') +
g.queryPanel('cortex_ingester_flush_queue_length{cluster="$cluster", job="$namespace/ingester"}', '{{pod}}'),
)
.addPanel(
g.panel('Flush Rate') +
g.qpsPanel('loki_ingester_chunk_age_seconds_count{cluster="$cluster", job="$namespace/ingester"}'),
),
)
.addRow(
g.row('Duration')
.addPanel(
g.panel('Chunk Duration hours (end-start)') +
g.queryPanel(
[
'histogram_quantile(0.5, sum(rate(loki_ingester_chunk_bounds_hours_bucket{cluster="$cluster", job="$namespace/ingester"}[5m])) by (le))',
'histogram_quantile(0.99, sum(rate(loki_ingester_chunk_bounds_hours_bucket{cluster="$cluster", job="$namespace/ingester"}[5m])) by (le))',
'sum(rate(loki_ingester_chunk_bounds_hours_sum{cluster="$cluster", job="$namespace/ingester"}[5m])) / sum(rate(loki_ingester_chunk_bounds_hours_count{cluster="$cluster", job="$namespace/ingester"}[5m]))',
],
[
'p50',
'p99',
'avg',
],
),
)
),
},
}
(import 'dashboards/loki-chunks.libsonnet') +
(import 'dashboards/loki-logs.libsonnet') +
(import 'dashboards/loki-operational.libsonnet') +
(import 'dashboards/loki-reads.libsonnet') +
(import 'dashboards/loki-writes.libsonnet')
172 changes: 172 additions & 0 deletions production/loki-mixin/dashboards/loki-chunks.libsonnet
Original file line number Diff line number Diff line change
@@ -0,0 +1,172 @@
local g = import 'grafana-builder/grafana.libsonnet';
local utils = import 'mixin-utils/utils.libsonnet';

{
grafanaDashboards+: {
local dashboards = self,

'loki-chunks.json':{
local cfg = self,

showMultiCluster:: true,
clusterLabel:: 'cluster',
clusterMatchers::
if cfg.showMultiCluster then
[utils.selector.re(cfg.clusterLabel, '$cluster')]
else
[],

namespaceType:: 'query',
namespaceQuery::
if cfg.showMultiCluster then
'kube_pod_container_info{cluster="$cluster", image=~".*loki.*"}'
else
'kube_pod_container_info{image=~".*loki.*"}',

assert (cfg.namespaceType == 'custom' || cfg.namespaceType == 'query') : "Only types 'query' and 'custom' are allowed for dashboard variable 'namespace'",

matchers:: {
ingester: [utils.selector.re('job', '($namespace)/ingester')],
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: I think this can be an equality match instead of a regexp.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I assume so, but I kept the same selectors as previously used for now. It is already a massive change :)

},

local selector(matcherId) =
std.join(',', ['%(label)s%(op)s"%(value)s"' % matcher for matcher in (cfg.clusterMatchers + cfg.matchers[matcherId])]),

ingesterSelector:: selector('ingester'),
ingesterSelectorOnly::
std.join(',', ['%(label)s%(op)s"%(value)s"' % matcher for matcher in cfg.matchers.ingester]),

templateLabels:: (
if cfg.showMultiCluster then [
{
variable:: 'cluster',
label:: cfg.clusterLabel,
query:: 'kube_pod_container_info{image=~".*loki.*"}',
type:: 'query'
},
] else []
) + [
{
variable:: 'namespace',
label:: 'namespace',
query:: cfg.namespaceQuery,
type:: cfg.namespaceType
},
],
} +
g.dashboard('Loki / Chunks')
.addRow(
g.row('Active Series / Chunks')
.addPanel(
g.panel('Series') +
g.queryPanel('sum(loki_ingester_memory_chunks{%s})' % dashboards['loki-chunks.json'].ingesterSelector, 'series'),
)
.addPanel(
g.panel('Chunks per series') +
g.queryPanel(
'sum(loki_ingester_memory_chunks{%s}) / sum(loki_ingester_memory_streams{%s})' % [
dashboards['loki-chunks.json'].ingesterSelector,
dashboards['loki-chunks.json'].ingesterSelectorOnly,
],
'chunks'
),
)
)
.addRow(
g.row('Flush Stats')
.addPanel(
g.panel('Utilization') +
g.latencyPanel('loki_ingester_chunk_utilization', '{%s}' % dashboards['loki-chunks.json'].ingesterSelector, multiplier='1') +
{ yaxes: g.yaxes('percentunit') },
)
.addPanel(
g.panel('Age') +
g.latencyPanel('loki_ingester_chunk_age_seconds', '{%s}' % dashboards['loki-chunks.json'].ingesterSelector),
),
)
.addRow(
g.row('Flush Stats')
.addPanel(
g.panel('Size') +
g.latencyPanel('loki_ingester_chunk_entries', '{%s}' % dashboards['loki-chunks.json'].ingesterSelector, multiplier='1') +
{ yaxes: g.yaxes('short') },
)
.addPanel(
g.panel('Entries') +
g.queryPanel(
'sum(rate(cortex_chunk_store_index_entries_per_chunk_sum{%s}[5m])) / sum(rate(cortex_chunk_store_index_entries_per_chunk_count{%s}[5m]))' % [
dashboards['loki-chunks.json'].ingesterSelector,
dashboards['loki-chunks.json'].ingesterSelector,
],
'entries'
),
),
)
.addRow(
g.row('Flush Stats')
.addPanel(
g.panel('Queue Length') +
g.queryPanel('cortex_ingester_flush_queue_length{%s}' % dashboards['loki-chunks.json'].ingesterSelector, '{{pod}}'),
)
.addPanel(
g.panel('Flush Rate') +
g.qpsPanel('loki_ingester_chunk_age_seconds_count{%s}' % dashboards['loki-chunks.json'].ingesterSelector,),
),
)
.addRow(
g.row('Duration')
.addPanel(
g.panel('Chunk Duration hours (end-start)') +
g.queryPanel(
[
'histogram_quantile(0.5, sum(rate(loki_ingester_chunk_bounds_hours_bucket{%s}[5m])) by (le))' % dashboards['loki-chunks.json'].ingesterSelector,
'histogram_quantile(0.99, sum(rate(loki_ingester_chunk_bounds_hours_bucket{%s}[5m])) by (le))' % dashboards['loki-chunks.json'].ingesterSelector,
'sum(rate(loki_ingester_chunk_bounds_hours_sum{%s}[5m])) / sum(rate(loki_ingester_chunk_bounds_hours_count{%s}[5m]))' % [
dashboards['loki-chunks.json'].ingesterSelector,
dashboards['loki-chunks.json'].ingesterSelector,
],
],
[
'p50',
'p99',
'avg',
],
),
)
){
templating+: {
list+: [
{
allValue: null,
current:
if l.type == 'custom' then {
text: l.query,
value: l.query,
} else {},
datasource: '$datasource',
hide: 0,
includeAll: false,
label: l.variable,
multi: false,
name: l.variable,
options: [],
query:
if l.type == 'query' then
'label_values(%s, %s)' % [l.query, l.label]
else
l.query,
refresh: 1,
regex: '',
sort: 2,
tagValuesQuery: '',
tags: [],
tagsQuery: '',
type: l.type,
useTags: false,
}
for l in dashboards['loki-chunks.json'].templateLabels
],
},
},
}
}
Loading