-
Notifications
You must be signed in to change notification settings - Fork 116
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
RS: v2 Prometheus metrics #540
base: release-rs-fuya-fuya
Are you sure you want to change the base?
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No idea about the PromQL, but language looks OK apart from some minor points you might want to consider. Also, one minor formatting glitch.
| <span class="break-all">bdb_other_req_max</span> | <span class="break-all">`sum by(bdb) (irate(endpoint_other_req[1m]))`</span> | Highest value of the rate of other (non read/write) requests on the database (ops/sec) | | ||
| <span class="break-all">bdb_other_res</span> | <span class="break-all">`sum by(bdb) (irate(endpoint_other_res[1m]))`</span> | Rate of other (non read/write) responses on the database (ops/sec) | | ||
| <span class="break-all">bdb_other_res_max</span> | <span class="break-all">`sum by(bdb) (irate(endpoint_other_res[1m]))`</span> | Highest value of the rate of other (non read/write) responses on the database (ops/sec) | | ||
| <span class="break-all">bdb_pubsub_channels</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Count the pub/sub channels with subscribed clients | |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
"Count" reads like an imperative here, maybe?
| <span class="break-all">bdb_pubsub_channels</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Count the pub/sub channels with subscribed clients | | |
| <span class="break-all">bdb_pubsub_channels</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Count of the pub/sub channels with subscribed clients | |
| <span class="break-all">bdb_other_res</span> | <span class="break-all">`sum by(bdb) (irate(endpoint_other_res[1m]))`</span> | Rate of other (non read/write) responses on the database (ops/sec) | | ||
| <span class="break-all">bdb_other_res_max</span> | <span class="break-all">`sum by(bdb) (irate(endpoint_other_res[1m]))`</span> | Highest value of the rate of other (non read/write) responses on the database (ops/sec) | | ||
| <span class="break-all">bdb_pubsub_channels</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Count the pub/sub channels with subscribed clients | | ||
| <span class="break-all">bdb_pubsub_channels_max</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Highest value of count the pub/sub channels with subscribed clients | |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Similar to the one above. There are a few more of these in the file.
| <span class="break-all">bdb_pubsub_channels_max</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Highest value of count the pub/sub channels with subscribed clients | | |
| <span class="break-all">bdb_pubsub_channels_max</span> | <span class="break-all">`sum by(bdb) (redis_server_pubsub_channels)`</span> | Highest value of count for the pub/sub channels with subscribed clients | |
| <span class="break-all">node_available_memory</span> | <span class="break-all">`node_available_memory_bytes`</span> | Amount of free memory in the node (bytes) that is available for database provisioning | | ||
| <span class="break-all">node_available_memory_no_overbooking</span> | <span class="break-all">`node_available_memory_no_overbooking_bytes`</span> | Available RAM in the node (bytes) without taking into account overbooking | | ||
| <span class="break-all">node_avg_latency</span> | <span class="break-all">`sum by (proxy) (irate(endpoint_acc_latency[1m])) / sum by (proxy) (irate(endpoint_total_started_res[1m]))`</span> | Average latency of requests handled by endpoints on the node in milliseconds; returned only when there is traffic | | ||
| <span class="break-all">node_bigstore_free</span> | <span class="break-all">`node_bigstore_free_bytes`</span> | Sum of free space of back-end flash (used by flash database's [BigRedis]) on all cluster nodes (bytes); returned only when BigRedis is enabled | |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not clear if this is supposed to be a plural or means something else.
| <span class="break-all">node_bigstore_free</span> | <span class="break-all">`node_bigstore_free_bytes`</span> | Sum of free space of back-end flash (used by flash database's [BigRedis]) on all cluster nodes (bytes); returned only when BigRedis is enabled | | |
| <span class="break-all">node_bigstore_free</span> | <span class="break-all">`node_bigstore_free_bytes`</span> | Sum of free space of back-end flash (used by flash databases [BigRedis]) on all cluster nodes (bytes); returned only when BigRedis is enabled | |
| --------- | :------------------- | :---------- | | ||
| <span class="break-all">listener_acc_latency</span> | <span class="break-all">N/A</span> | Accumulative latency (sum of the latencies) of all types of commands on the database. For the average latency, divide this value by listener_total_res | | ||
| <span class="break-all">listener_acc_latency_max</span> | <span class="break-all">N/A</span> | Highest value of accumulative latency of all types of commands on the database | | ||
| <span class="break-all">listener_acc_other_latency</span> | <span class="break-all">N/A</span> | Accumulative latency (sum of the latencies) of commands that are a type "other" on the database. For the average latency, divide this value by listener_other_res | |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I guess "a type XXX" is OK for these but "of type XXX" is used elsewhere in the table (not a major problem!)
| <span class="break-all">redis_keyspace_write_misses</span> | <span class="break-all">`redis_server_keyspace_write_misses`</span> | Number of write operations accessing a non-existing keyspace | | ||
| <span class="break-all">redis_master_link_status</span> | <span class="break-all">`redis_server_master_link_status`</span> | Indicates if the replica is connected to its master | | ||
| <span class="break-all">redis_master_repl_offset</span> | <span class="break-all">`redis_server_master_repl_offset`</span> | Number of bytes sent to replicas by the shard; calculate the throughput for a time period by comparing the value at different times | | ||
| <span class="break-all">redis_master_sync_in_progress</span> | <span class="break-all">`redis_server_master_sync_in_progress`</span> | The master shard is synchronizing (1 true | 0 false) | |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks like the vertical bar here creates an extra invisible column, so you can't see the "0 false" part.
| <span class="break-all">redis_master_sync_in_progress</span> | <span class="break-all">`redis_server_master_sync_in_progress`</span> | The master shard is synchronizing (1 true | 0 false) | | |
| <span class="break-all">redis_master_sync_in_progress</span> | <span class="break-all">`redis_server_master_sync_in_progress`</span> | The master shard is synchronizing (1 true; 0 false) | |
… instead of PromQL on dedicated v2 metrics page
DOC-3944