Ops Tools
225 results found
-
Ops Manager API should be authenticated via Certificate
Ops Manager API uses HTTP Digest Authentication, this method requires a username and a password which are hashed and provided in the form of registered Ops Manager account, and the password which is a Public API Key associated with that account.
But as part of security feature we should be allowed to use Authentication based on SSL certificates as we are using for Client connection, This will help to secure the self automation process if we are developing any internally.
3 votes -
When changing snapshot retention, prompt user to apply new policy to existing snapshots
Currently when you change the snapshot schedule, the changes only apply to NEW snapshots.
Please allow the user the option to apply the new policy automatically to all existing snapshots. Or at the very least, notify them of them of snapshots that exist which do not meet the current policy!
Let me provide an example of the problem. If you have a retention policy of 10 days and on DAY1 you change it to 30 days, your retention will look like this:
DAY1 - 10 snapshots
DAY2 - 10 snapshots
...
DAY10 - 10 snapshots
DAY11 - 11 snapshots
DAY12…1 vote -
Lock enableLocalConfigurationServer setting on OPS Manager side
To harden security for mongodb deployment managed by OPS Manager, we can use setting enableLocalConfigurationServer = true so automation-mongod.conf won't have any passwords for ssl certs and agent will retrieve them from OPS Manager.
to disable this feature and to read all passwords for ssl certs (and hence get access to mongodb data) it's enough to comment out this parameter an restart automation service (or wait until host will be restarted).
Linux root user can modify any file on mongodb host including this file and can restart any services, so it's impossible to protect getting all passwords and mongod.conf from…
1 vote -
The Ops Manager UI provide an option to trigger election (stepDown)
Have the option to perform primary stepDown from Ops Manager UI to trigger election
4 votes -
I found what I believe to be an error in the following document
I found what I believe to be an error in the following document.
https://www.mongodb.com/docs/ops-manager/current/tutorial/install-simple-test-deployment/In the section "5. Create the Ops Manager Application Database directory," it instructs to execute the following command
sudo chown -R mongod:mongod /dataIn my environment, the mongod user does not exist and I get an error, but looking at the passwd file, the mongodb user and mongodb group seem to exist. The version is 11.7.
I think the correct command is as follows
sudo chown -R mongodb:mongodb /data1 vote -
Operating System distribution and version of a host in OPS Manager API
Hi,
would be useful having the operating system distribution and version of a host for our automation scripts.
This info is not available in any OPS Manager API request, as the case 01119828.
My suggestion is add this info at "Get Host by ID" https://www.mongodb.com/docs/ops-manager/current/reference/api/hosts/get-one-host-by-id/
Best regards,
Danilo1 vote -
Ability to mark a deployment as an INELIGIBLE restore target
Restoring to a cluster is one of the few destructive actions that Ops Manager takes and it's terrifying to see our main production cluster listed as a possible restore target!
I would love to be able to toggle a setting on this cluster to indicate that it is NOT available as a restore target.
This could be similar to the AWS "DisableApiTermination" feature that prevents instance termination.
https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/terminating-instances.html#Using_ChangingDisableAPITermination1 vote -
Ops Manager Load Balancer
In the same way we can connect to a replica set via a connection string containing the 3 hostnames of the members (in a 3 member replica set) which indicates some load balancing built in, it would be good to allow automation agent's mmsBaseUrl parameter to accept 3 Ops Manager web app hostnames if Ops Manager was running in 3 nodes on top of a 3 member replica set OM Application DB. With this, if one of the 3 Ops Manager web apps fails, the agent's connection to OM will be uninterrupted. Currently we will have to configure our own…
2 votes -
Deploy Changes without restarting mongod/mongos instance immediately.
Whenever we want to make changes, eg. set a new parameter or add new parameter in configuration (advance configuration options), after we save changes, review and deploy, automation immediately starts applying that change and does a rolling restart.
We need flexibility in restart, means one should have an option to perform immediate rolling restart or defer it to later time. We may apply multiple changes at different times and set one preferred window to restart instance instead of doing multiple restarts.16 votes -
Add recommended ulimit settings to the systemctl service definition
The Ops Manager documentation recommends higher than default ulimit settings, so why not include these in the systemctl service definition?
2 votes -
Ability to start the bi connector using a .drdl file in Ops Manager
Ability to start the bi connector using a .drdl file in Ops Manager. Currently only possible with en premise bi connector deployment(mongosqld).
1 vote -
Add possibility to configure the process hostnames for the automation config map
In order to use external certificates could it be possible to configure the process hostnames in automation config map
1 vote -
TLS secret key config
Add option to configure the secret keys for the server and CA certificate and replace the default ones tls.crt, tls.key and ca.crt
1 vote -
Add Ops Manager alert on tlsX509ExpirationWarningThresholdDays client
Ops Manager alert, for messages posted through 'tlsX509ExpirationWarningThresholdDays' parameter to warn for client certs expiry
1 vote -
Disk metric - log drive
The ops manager is not capturing the log drive disk related metrics when the log path is different from data path. This is an important metric to monitor for anyone to consider OM as the monitoring tool.
1 vote -
add logical name to disk in metrics
when we look at hardware metric in replicaset, OPS Manager displays and groups disk stats based on internal disk name on each node. if for some reason we have disks discovered in different order on OS level, disk names will be different and it's become very messy and difficult to compare disk stats between different nodes.
Suggestion is to add ability to give disk logical name so OPS Manager will display disks for data and disks for logs in a nice way. if new disk is replaced or added - default name can be displayed so administrator will be able…
1 vote -
Add milliseconds to Ops Manager access logs
The current logback-access.xml file only logs with second grainularity, which isn't great when trying to correlate HTTP events with other log events.
I propose changing this line:
<pattern>%h %l %u [%t] "%X{requestUrl}" %s %b "%i{Referer}" "%i{User-Agent}"
To this:
<pattern>%h %l %u [%date{"dd/MMM/yyyy:HH:mm:ss.SSS Z"}] "%X{requestUrl}" %s %b "%i{Referer}" "%i{User-Agent}"
1 vote -
Option to include hidden secondaries in Replica Set Alerts
As per Replica Set Alerts documentation: "A replica set member is healthy if you run rs.status() for that replica set and the result returns PRIMARY or SECONDARY for that member. Hidden secondaries and arbiters are not counted".
That is convenient when you take hidden secondaries down for backups but we would like include the hidden members in replica set health monitoring.
2 votes -
trigger a queryable Backup Snapshot restore via Ops Manager public API
We would like to periodically provision queryable snapshots on a programmatic basis (aka Ops Manager public API call) to easily get a consistent dump of sharded clusters in order to put those dumps in a "very safe location". We want to use this as part of a last resort backup strategy. The idea is that a dump is much easier to be restored by a "last resort restore app" compared to the http-downloaded plain data files.
One other option to extract the contents of a sharded cluster would be to use "Long Running Queries" but this cannot be used along…
5 votes -
1 vote
- Don't see your idea?