Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Browse files

Separate replication ratio with repliation number

  • Loading branch information...
commit 321fbce5e27bf57f46cfeda6bab66389410609e1 1 parent 145dbc7
@bcui6611 bcui6611 authored
Showing with 29 additions and 7 deletions.
  1. +23 −6 diskqueue_stats.py
  2. +6 −1 prescription.py
View
29 diskqueue_stats.py
@@ -222,11 +222,11 @@ def run(self, accessor, scale, threshold=None):
"perBucket" : True,
"perNode" : True,
},
- {"name" : "ReplicationTrend",
+ {"name" : "ReplicationPercentageTrend",
"ingredients" : [
{
- "name" : "replicationTrend",
- "description" : "Replication backlog size or replication ratio",
+ "name" : "replicationPercentageTrend",
+ "description" : "Replication backlog size to active item ratio",
"counter" : ["curr_items", "ep_tap_total_total_backlog_size"],
"scale" : "hour",
"code" : "ReplicationTrend",
@@ -235,19 +235,36 @@ def run(self, accessor, scale, threshold=None):
"low" : 10.0,
"high" : 30.0,
},
+ },
+ "symptom" : "Number of backlog item to active item ratio '{0}%' is above threshold '{1}%'",
+ "formula" : "Avg(ep_tap_total_total_backlog_size) / Avg(curr_items) > threshold",
+ }
+ ],
+ "perBucket" : True,
+ "indicator" : True,
+ },
+ {"name" : "ReplicationNumTrend",
+ "ingredients" : [
+ {
+ "name" : "replicationNumTrend",
+ "description" : "Replication backlog size",
+ "counter" : ["curr_items", "ep_tap_total_total_backlog_size"],
+ "scale" : "hour",
+ "code" : "ReplicationTrend",
+ "threshold" : {
"number" : {
"low" : 50000,
"high" : 100000,
},
},
- "symptom" : "Number of backlog items or replication ratio '{0}%' is above threshold '{1}%'",
- "formula" : "Avg(ep_tap_total_total_backlog_size) / Avg(curr_items) > threshold",
+ "symptom" : "Number of backlog items '{0}%' is above threshold '{1}%'",
+ "formula" : "Avg(ep_tap_total_total_backlog_size) > threshold",
}
],
"perBucket" : True,
"indicator" : True,
},
- {"name" : "DiskQueueDrainingAnalysis",
+ {"name" : "DiskQueueDrainingAnalysis",
"description" : "",
"ingredients" : [
{
View
7 prescription.py
@@ -52,11 +52,16 @@
"impact" : "Data is available in memory but any data in the queue waiting to be persisted will be lost if the node goes down",
"action" : "Increase disk quota for buckets, or add nodes to cluster. Can remove and re-add a server to resolve a disk fragmentation issue. If issue persists please contact support@couchbase.com"
},
- "ReplicationTrend" : {
+ "ReplicationNumTrend" : {
"cause" : "Replication queue overloaded",
"impact" : "If the nodes fails over, data will be missing on the replica. If you failover, information in the replication queue will be lost.",
"action" : "Do not failover the node. Wait until replication queue is low to failover. If replication queue remains high, contact support@couchbase.com",
},
+ "ReplicationPercentageTrend" : {
+ "cause" : "Replication severely behind",
+ "impact" : "If the nodes fails over, data will be missing on the replica. If you failover, information in the replication queue will be lost.",
+ "action" : "Do not failover the node. Wait until replication queue is low to failover. If replication queue remains high, contact support@couchbase.com",
+ },
"DiskQueueDrainingAnalysis" :{
"cause" : "Drain rate from RAM to disk is too slow; can be caused by disk fragmentation.",
"impact" : "Rate of data persisted from RAM to disk is too high.",
Please sign in to comment.
Something went wrong with that request. Please try again.