We have discovered an issue in the new 6.x versions of check_netapp_pro if a getter contacts the new RESTful API of NetApp. These getters require an explicit --stm (e.g. --stm=1h). Otherwise the collected storefiles are not deleted and could fill up the monitoring servers disk.
Again this does affect installations only if both of the following conditions are met:
the plugins version is 6.x the plugins getter contacts the RESTful API of the filer (Ontap 9.
We have at least one confirmed report, that our latest release v6.0.0 has performance problems in large environments, when running many checks at the same time.
There seems to be a limit of approximately 12 checks per second due to the overhead of running a docker container. We are working on a solution and will report here once we can provide an updated version.
Customers with large environments may wait before upgrading to v6.
After extensive testing we released the next major version 6.0.0 today. As already announced this version brings:
Easy installation and update without dependencies on Perl or Perl-Modules. Getter amd checks compatible with NetApps new RESTful API New checks and options More details are in the release history.
This version is already available on your distributors portal. The completely rewritten and new documentation is available online at docs.monitoring-plugins.pro.
We just pushed the first RC for the upcoming 6.0.0 version to the Distributors Portal. Customers who would like to test are very much appreciated! (Pls. check with the distributor to get access to unstable releases as well, if you do not see them on your account.) The 6.0.0 version brings a completely dockered and therefore dependency free installation plus new checks and options. Please read the release history for details.
Since 4.0.0 every over-all check has a feature, which we call “instance-name contains node-name". We have introduced this as an replacement for the sporadic available --node|--vserver parameter (not available any more). So since then each instance-name has the name of its node prefixed. This allows to use the --include|--exclude parameters to filter for all instances of a specific node.
E.g. if you have a new-node with a different hardware you may want to check only the ports of this new-node:
We are constantly trying new methods of installing the plugins so to make installation and upgrade easy and scriptable (rpm, …). We found the Perl dependencies being the main obstacle and finally favorite a solution depending on docker.
So in the future you could get a fully installed docker image with anything inside and ready to run. This image will of course require docker to run on your monitoring server. But in turn you won’t need to have Perl or any of its plugins installed.
Our quite new SnapCenter check can filter for resource-groups and policies. Until now one had to precisely define which resource-groups should get checked.
In some situations the resource-group name changes. To allow checking such resource-groups without having to constantly change the filters in the service-check configuration we are introducing regexmatching for these filters.
Example Let me show you an example:
If a resource-group sometimes changes from …
AT01_HA_D01 to AT01_HA_D01_SR (and vice versa),
We have one report, that changing the setup from a chain setup (netapp01 → netapp02 → netapp03) to a star setup (netapp01 → netapp02 ➕ netapp01 → netapp03) makes stoping some collectors (e.g. get_netapp_cm.pl) with an error of “Instance ‘NETAPP01-SVM01:NETAPP01_SVM01_xxxx01_vol’ already exists - can not continue!”
If you see this error in your setup please get in contact with us.
We are happy to announce that the next major release of check_netapp_pro is now available on the Q-Portal for download.
The breaking changes for this release are:
A new directory layout which will ease the creation of RPM-packages. Also we have introduced a new getter for DataONTAP 9.7s new API plus several new checks. Some of these are already available as a compiled binary with zero dependencies. Several of the getters and checks now require a license-file, which you will find at the Q-Portal.
The new UnprotectedVolume check searches for volumes not protected by a SnapMirror.
An alarm is generated if one or more unprotected volumes are found. Specific volumes and vservers can be checked by using the standard check_netapp_pro --include= and --exclude= arguments. A simple example follows using the default arguments and a single host
./check\_netapp\_pro.pl UnprotectedVolume -H filer NETAPP\_PRO UNPROTECTEDVOLUME CRITICAL - 7 volumes checked, 6 critical and 0 warning mycluster-01:vol0: unprotected (CRITICAL) vserv\_a:vol0: unprotected (CRITICAL) vserv\_b:vol0: unprotected (CRITICAL) vserv\_b:vol1: unprotected (CRITICAL) vserv\_b:vol1\_mirror: unprotected (CRITICAL) vserv\_b:vol2: unprotected (CRITICAL) vserv\_a:vol1: protected (vserv\_b:vol1\_mirror, data\_protection) This check is part of the upcoming 5.