Alerts (MongoDB)

The Management Pack for MongoDB creates alerts (and in some cases provides recommended actions) based on various symptoms it detects in your MongoDB Environment. See the table below for the list of alerts available in the Management Pack.

Alerts List

Name	Description	Recommendation
Replica Set has even number of voters	This alert indicates that an Immediate alert was raised in MongoDB.	A replica set should always have an odd number of members so there is never a tie when voting for the primary host. Either add another replicated host, or add an arbiter to the replica set.
Replica Set Has Less than 3 Data Bearing Nodes	This alert indicates that an Immediate alert was raised in MongoDB.	Three data-bearing nodes are required for replica-set wide data durability.
Mongod Shares a Host	This alert indicates that an Immediate alert was raised in MongoDB.	Running multiple mongod instances on a single host increases the likelihood of having inaccessible data by reducing horizontal scaling.
Mongod Is Recovering From Delay	This alert indicates that an Immediate alert was raised in MongoDB.	This instance is in RECOVERING mode.
Mongod Has a Delay Behind the Primary	This alert indicates that an Immediate alert was raised in MongoDB.	This node has very high replication delays. Delayed nodes can cause out-of-date data to be presented.
Mongod Has a Delay Behind the Primary	This alert indicates that an Immediate alert was raised in MongoDB.	This node has very high replication delays. Delayed nodes can cause out-of-date data to be presented.
Mongod Journaling Disabled	This alert indicates that an Immediate alert was raised in MongoDB.	To get accurate snapshots of running mongod processes, journaling must be enabled. This is critical if using the MMAPv1 storage engine.
Mongos Total Connections Reaching Maxmimum	This alert indicates that an Immediate alert was raised in MongoDB.	Current connections approaching max connections.
Mongos Ratio of Page-Faults to Operations Greater than or Equal to 1	This alert indicates that an Immediate alert was raised in MongoDB.	There are a large number of page faults. Consider increasing memory provisioning.
Mongos is Reaching a Congested Level of Connections	This alert indicates that an Immediate alert was raised in MongoDB.	Number of connections has greatly increased compared to normal. Large numbers of connections use large amounts of memory, so be sure to provision adequate memory on the host.
Config Server Has no Configured Replica Set	This alert indicates that an Immediate alert was raised in MongoDB.	A production config server should always be configured as a replica set; when configured as a single host, if that host goes down, the entire cluster becomes inaccessible.
Shard Has No Configured Replica Set	This alert indicates that an Immediate alert was raised in MongoDB.	If a shard is not configured with a replica set, then if the host it is configured with goes down, none of the data on that host will be available. Configuring the shard as a replica set provides better fallback, allowing for a system with higher-availability.