I recently had a customer that needed to have the overall health, utilization, performance, and any system events reported to a central location as their environment was spread out over a large number of clusters in varied geographies. After discussing the use of SNMP to meet this need they asked for an explanation of what types of information they would be capable of receiving.
I explained to them that all of this information was available in our MIB files, which can be downloaded from the Isilon UI.
These MIBs can be viewed by using MIB Browsers and contain an extensive amount of information about the available traps that can be sent from Isilon clusters. There are many available MIB Browsers available for free online, and can be found for any operating system, with a quick search for "MIB Browser" or "MIB Walker" using your favorite search engine.
I also included a brief summary of that data in each MIB to the customer, which I have included below, but still recommended a MIB walk for the most in depth information.
Provides information at the Cluster and Node levels
The Cluster Status portion contains information such as the Cluster Name, Overall Health, GUID, Node Count, the number of configured nodes and which are nodes in the cluster are online or offline.
The Cluster Performance portion include IFS and Cluster Network performance information, such as bytes in and out and overall throughput. You will also find information related to the Cluster CPU performance that includes CPU time for a variety of processes as well as idle time.
The IFS File System portion of this MIB includes information related to the Total Bytes in the Cluster as well as bytes Used and Available.
Licenses includes a license table with information on what Cluster licenses are being used, or not used, on this cluster and whether they are licensed, unlicensed, or an evaluation license.
Quotas includes a table of specific directory, user, and group quotas. This includes information on quoits type, ID, path, whether hard or soft thresholds are defined, and of course, usage.
The Snapshots portion of this MIB includes information in regard to if automatic snapshot creation or deletion is enabled, what percentage of storage is reserved for snapshots, and whether root or subdirectories are visible to Local, NFS, or CIFS clients. Schedule and Snapshot tables are also included which include data on what snapshots are scheduled for creation or expiration and what path, aliases, and locks are in place.
Node Status includes node name, health, type, and whether or not the node is in read-only mode.
Node Performance includes information at the node level similar to the cluster performance stats referenced above. However, at the node level the addition of tables providing Node Protocol information relating operation count, throughput, and latency and a Disk Performance table including device names, operations and input and output rates per second are added.
The Node portion of this MIB also contains several tables such as the Chassis, Disk, Fan, TempSensor, and PowerSensor tables. These tables include corresponding names logical numbers, model numbers, firmware versions, and current statuses of these areas.
Provides information on all available Isilon cluster and node traps.
This MIB contains all of the traps that can be sent via SNMP from the Isilon Cluster and Nodes. This MIB is very extensive and covers entirely too many traps to describe with any type of brevity, however I will attempt a high level description below.
Group State Traps contains node traps that report node status, expansion node status, and whether the node is being shutdown, rebooted, or if either of these actions has failed.
File System Traps contain traps related to partition usage, overall cluster and disk pool usage, and whether and file system errors or warnings have occurred.
There are multiple groups of traps for various areas of cluster and node function that will often not report specific utilization numbers or percentages, but will report errors and warnings on the devices, such as whether a threshold has been reached or exceeded, whether the notification of this state has succeeded, and whether a device is functioning as expected. These areas are briefly listed below:
Smart Quotas - Threshold, Notification and Error warnings
Snapshot - Error and Reserve warnings
SyncIQ - Event, Policy and Target Errors
Software - Numerous traps covering Job Engine, Flex Protect, AVScan Infected Files, MCP Process failures, license expiration, firmware updates, test event, Node CPU, Client Mount and Host Lookup failures.
AVScan - Detects whether ICAP servers have been configured and reports Scan errors
NetProtocols - Contains traps relevant to ADS clock skew, AD Connectivity, Win ID Map range, and various Windows authentication components.
Networking - Reports network interface or overall network problems.
Hardware - Contains traps related to hardware, sensor, fan, temperature, and voltage.
System Disk - Drive status and disk errors
Sensor - Contains Power supply, Fan, Voltage, Temperature, Power, Memory, CPU, and Current information.
Storage Transport - SAS PHY and Storage Transport information.
Again, this is a brief overview of what is contained in each of these MIBs. More information about each leaf and trap can be viewed by using a MIB Browser and selecting the corresponding item.
Our customer has Insight IQ setup for monitoring and using the Isilon "E-mail notifications" for notifications regarding events and alerts. Now they are asking to setup the SNMP for both Monitoring and setting alert traps. Is this advisable?
Thanks for the info on SNMP.
Our tools team (bmc truesight operations management) using to discovery any devices in the network so they asked to provided Ision MIB in the .mib or exe format .But we can only get MIB files in txt format. Do you have any insights on this MIB format issue.
you can ask Isilon support to provide the corresponding .mib file. But try in google and you may get a link.
Also there is a mib compiler free link in the internet. You can run the compiler test and validate the file you have or any dependent mib that may be needed.