ZooKeeper by Prometheus 139,841139,841 3.5 (2 reviews)
- Tested with: the ZooKeeper new metric system since 3.6.0
- Please contribute, comment, feedback, suggestions here or write an email to maoling199210191@sina.com.
- If you have any issue, you can also report it to https://issues.apache.org/jira/projects/ZOOKEEPER/issues, label it as metric system
- Users can ask for this Grafana dashboard account if having any good improvements by writing an email to dev@zookeeper.apache.org.
- Installation guide can be found: https://github.com/apache/zookeeper/blob/master/zookeeper-docs/src/main/resources/markdown/zookeeperMonitor.md
- For Alerting with Prometheus, Look at: https://github.com/apache/zookeeper/blob/master/zookeeper-docs/src/main/resources/markdown/zookeeperMonitor.md#Alerting
Change Log
[v5]
- TODO: Adapt to kubernetes(k8s) and container environment
[v4]
Feature:
- Add the JVM related metrics
- Support to choose by datasource, job and instance
Fix:
- Remove duplicated metrics: e.g., write_batch_time_in_commit_processor
Improvement:
- Use rate() to react the fluctuation of data for total time summary metric
- Have a better classification by checking the source codes to know what every metric wants to do
- The metrics in the CommitProcessor are placed the same order with the source code for a better observation for the workflow of CommitProcessor
- Use an unit(e.g., ms/byte) for metric graph
- Prioritize the metrics in the same panel
Compatibility Test: - Tested with grafana-6.x(6.2.2); grafana-7.x(7.0.5)
Used Metrics 8686
znode_count
ephemerals_count
global_sessions
local_sessions
write_per_namespace_sum
read_per_namespace_sum
approximate_data_size
packets_received
packets_sent
response_packet_cache_misses
response_packet_cache_hits
response_packet_get_children_cache_misses
response_packet_get_children_cache_hits
unrecoverable_error_count
digest_mismatches_count
startup_snap_load_time
startup_snap_load_time_count
startup_snap_load_time_sum
startup_txns_loaded
startup_txns_loaded_count
startup_txns_loaded_sum
dbinittime
dbinittime_count
dbinittime_sum
open_file_descriptor_count
fsynctime
fsynctime_count
fsynctime_sum
snapshottime_sum
snapshottime_count
prep_process_time_sum
prep_processor_queue_time_ms_sum
prep_processor_queue_size_sum
prep_processor_request_queued
outstanding_changes_queued
outstanding_changes_removed
close_session_prep_time_sum
sync_process_time_sum
sync_processor_queue_flush_time_ms_sum
sync_processor_queue_size_sum
sync_processor_request_queued
sync_processor_batch_size_sum
commit_process_time_sum
read_commitproc_time_ms_sum
write_commitproc_time_ms_sum
write_commit_proc_req_queued_sum
read_commit_proc_req_queued_sum
commit_commit_proc_req_queued_sum
requests_in_session_queue_sum
read_commit_proc_issued_sum
write_commit_proc_issued_sum
concurrent_request_processing_in_commit_processor_sum
time_waiting_empty_pool_in_commit_processor_read_ms_sum
pending_session_queue_size_sum
local_write_committed_time_ms_sum
server_write_committed_time_ms_sum
write_batch_time_in_commit_processor_sum
reads_after_write_in_session_queue_sum
session_queues_drained_sum
reads_issued_from_session_queue_sum
request_commit_queued
write_final_proc_time_ms_sum
read_final_proc_time_ms_sum
readlatency_sum
updatelatency_sum
max_latency
min_latency
avg_latency
proposal_latency_sum
quorum_ack_latency_sum
ack_latency_sum
propagation_latency_sum
commit_propagation_latency_sum
proposal_ack_creation_latency
tls_handshake_exceeded
outstanding_tls_handshake
ensemble_auth_fail
ensemble_auth_success
ensemble_auth_skip
jvm_classes_loaded
jvm_threads_current
jvm_threads_deadlocked
jvm_pause_time_ms_sum
jvm_gc_collection_seconds_sum
jvm_threads_state
jvm_memory_pool_bytes_used