What are the Polling interval best practices?

FormerMember · 2016-03-11T18:26:38+00:00

Polling Intervals Default Node Poll Interval seconds Check response time and status every configured time interval. Default Interface Poll Interval seconds Check response time and status every configured time interval. Default Volume Poll Interval seconds Check response time and status every configured time interval.…

FormerMember

Polling Intervals
Default Node Poll Interval	seconds	Check response time and status every configured time interval.
Default Interface Poll Interval	seconds	Check response time and status every configured time interval.
Default Volume Poll Interval	seconds	Check response time and status every configured time interval.
Default Rediscovery Interval	minutes	Default rediscovery interval.
Default Asset Inventory Poll Interval	days	Check asset inventory every configured time interval

Here are my current polling interval for about 1100 nodes. Does this seem reasonable? I believe that these are the defaults since I clicked on the "reapply polling intervals" button in the polling settings page.

Find more posts tagged with

Performance

npm

Polling

Accepted answers

All comments

bspencer63

ikomrad,

Have you checked out this thread?

Re: Poller Tuning - Community Tips

No matter the polling interval you choose, you need to ensure your polling engine is completing polls before the next poll starts.

Hope this helps...

BSpencer

FormerMember

I did but it's way over my head. I think I need straightforward instructions such as

log in
click here
click here
record this number
record another number
plug them into this formula
take your answer and enter it in this field
click on submit for your change to take effect.

bspencer63

ikomrad,

First off I would ask what are you trying to accomplish by changing the defaults?

Changing the defaults will affect your NPM server drastically depending upon which way you adjust the polling interval. From the default of 120 seconds to 60 seconds can actually make your server unreachable if you are polling a large number of nodes and interfaces. Decreasing from 120 might not keep you up with current status.

To change the polling interval on all devices:

Checking and Changing Polling Cycles and Data Retention - YouTube

View the below YouTube video for directions and advice. "SolarWinds Head Geeks Leon Adato, Kong Yang, and Patrick Hubbard dig into ways to verify both the polling timing: how often monitoring polls are performed, as well as data retention settings that might bog down your monitoring solution if not correctly configured."

But if that isn't enough, here is what you do:

Select

NPM Home Page

Then Select

Admin

Then Select

Thresholds & Polling (This is where the settings are adjusted) ( Configure polling settings to optimize polling performance )

On the above page you will find the way to set all polling times. This will adjust them on all, I repeat ALL assigned and monitored nodes, interfaces, volumes, etc.

I would suggest you read the below articles and learn from others mistakes by searching Thwack for proper tweaking and alerting to ensure you do not overwork your SW Server.

Best Practice could be to just accept the default polling intervals, but I found that in some instances defaults just will not complete what I need to monitor. Some interfaces I need to poll more frequently and some I just don't care about. This is why I stressed above that to change the polling is an all or nothing from the "Thresholds & Polling" Admin page. You are able to tweak them on a per node/interface/volume etc from the node Details, Edit pages or from the DB itself, that is if you are confident enough to make changes to your DB. If not, you are at the blanket change state. Not terrible, but it will either add or diminish pressure on your NPM server.

Re: Polling intervals tunning ? ?

Re: NPM Poller Information

Re: Status Polling

SolarWinds Knowledge Base :: Orion Poller Configuration Rebuild/Migration

SolarWinds Knowledge Base :: Poller Status Warning: A poller is either approaching or in excess of its polling rate limi…

Default Polling Settings from the online Admin Guide are:

Polling Settings

The Orion Poller Settings view establishes both the polling intervals that are observed by the Orion NPM poller and the retention periods maintained for the Orion database. The following Orion NPM poller settings are configured diirectly from the Orion Web Console.

Note: Changes to database maintenance and retention settings do not take effect until the SolarWinds Network Performance Monitor service is restarted.

Polling Intervals

The following settings configure the intervals on which Orion completes ICMP status and response time polls.

Default Node Poll Interval

Devices are regularly polled to determine status and response time on the period of time designated as the Default Node Poll Interval. By default, this interval is 120 seconds.

Default Interface Poll Interval (Orion NPM)

If Orion NPM is installed, interfaces are regularly polled to determine status and response time on the period of time designated as the Default Interface Poll Interval. By default, this interval is 120 seconds.

Default Volume Poll Interval

Volumes are regularly polled to determine status and response time on the period of time designated as the Default Volume Poll Interval. By default, this interval is 120 seconds.

Default Node Topology Poll Interval

Monitored devices are regularly polled for topology data on this interval. By default, this interval is 30 minutes.

Default Rediscovery Interval

Your entire network is polled on this interval to detect any re‑indexed interfaces. Monitored network devices are also checked for IOS upgrades permitting EnergyWise support. By default, this interval is 30 minutes.

Note: In Orion products released prior to Orion NPM version 10.1, the minimum interval allowed is 1 minute. Beginning with Orion NPM version 10.1, the minimum rediscovery interval is 5 minutes. Polling interval settings may not be submitted if the default rediscovery interval is not set to at least 5 minutes.

Lock Custom Values

Checking this option stores the custom ICMP polling interval settings you have configured.

Polling Statistics Intervals

The following settings configure the intervals on which Orion completes SNMP statistics polls.

Default Node Statistics Poll Interval

Device performance statistics are regularly polled on the period of time designated as the Default Node Statistics Poll Interval. By default, this interval is 10 minutes.

Default Interface Statistics Poll Interval

If Orion NPM is installed, interface performance statistics are regularly polled on the period of time designated as the Default Interface Statistics Poll Interval. By default, this interval is 10 minutes.

Default Volume Statistics Poll Interval

Volume performance statistics are regularly polled on the period of time designated as the Default Volume Statistics Poll Interval. By default, this interval is 15 minutes.

Dynamic IP Address and Hostname Resolution

Select the Default IP Address Resolution (IPv4 or IPv6) for dual stack objects on your network.

Database Settings

The following settings configure the database maintenance archive creation time and retention periods for selected data in your Orion database. Additionally, the alert timeout (Max Alert Execution Time) and the text displayed to prompt for web-based alert acknowledgement (Alert Acknowledgment URL Text) are included.

Note: It may take upwards of 10 minutes to fully propagate some changes to Orion database settings.

Archive Time

The Archive Time is the time of the day when maintenance of the Orion database occurs. For more information about database maintenance, see “Database Maintenance” in the SolarWinds Orion Network Performance Monitor Administrator Guide.

Detailed Statistics Retention

All statistics in the Orion database that are collected on any basis shorter than 1 hour are summarized into hourly statistics after the period of time designated as the Detailed Statistics Retention period. By default, this period is 7 days.

Hourly Statistics Retention

All statistics in the Orion database that are collected on any basis shorter than 1 day but longer than 1 hour are summarized into daily statistics after the period of time designated as the Hourly Statistics Retention period. By default, this period is 30 days.

Daily Statistics Retention

All statistics in the Orion database that are collected on a daily basis are kept for the period of time designated as the Daily Statistics Retention period, and then they are deleted. By default, this period is 365 days.

Detailed Interface Availability Statistics Retention

All interface availability statistics in the Orion database that are collected on any basis shorter than 1 hour are summarized into hourly statistics after the period of time designated as the Detailed Interface Availability Statistics Retention period. By default, this period is 7 days.

Hourly Interface Availability Statistics Retention

All interface availability statistics in the Orion database that are collected on any basis shorter than 1 day but longer than 1 hour are summarized into daily statistics after the period of time designated as the Hourly Interface Availability Statistics Retention period. By default, this period is 30 days.

Daily Interface Availability Statistics Retention

All interface availability statistics in the Orion database that are collected on a daily basis are kept for the period of time designated as the Daily Interface Availability Statistics Retention period, and then they are deleted. By default, this period is 365 days.

Detailed Wireless Statistics Retention

All wireless statistics in the Orion database that are collected on any basis shorter than 1 hour are summarized into hourly statistics after the period of time designated as the Detailed Wireless Statistics Retention period. By default, this period is 7 days.

Hourly Wireless Statistics Retention

All wireless statistics in the Orion database that are collected on any basis shorter than 1 day but longer than 1 hour are summarized into daily statistics after the period of time designated as the Hourly Wireless Statistics Retention period. By default, this period is 30 days.

Daily Wireless Statistics Retention

All wireless statistics in the Orion database that are collected on a daily basis are kept for the period of time designated as the Daily Wireless Statistics Retention period, and then they are deleted. By default, this period is 365 days.

Events Retention

All network events data is deleted from the Orion database after the period of time designated by the Events Retention has passed after the event ending time. By default, this period is 30 days.

Syslog Messages Retention

All network data related to received Syslog messages is deleted from the Orion database after the period of time designated by the Syslog Messages Retention, starting when the message is received, has passed. By default, this period is 7 days.

Trap Messages Retention

All network data related to received trap messages is deleted from the Orion database after the period of time designated by the Trap Messages Retention, starting when Orion receives the trap message, has passed. By default, this period is 30 days.

Max Alert Execution Time

If an alert takes longer than the period of time designated by this setting, the alert is disabled. If the defined alert condition persists, Orion will continue trying to execute the alert.

Alert Acknowledge URL Text

The text provided in this field displays when alerts are available for acknowledgment over the web. When viewing an alert, click the provided text to acknowledge the viewed alert.

Allow alert actions for unmanaged objects

Check this option if you want the SolarWinds Alerting Engine to execute configured alert actions for unmanaged objects. Enabling this option will increase the processing load on both your SolarWinds server and your database server.

Discovery Retention

All network discovery profiles are deleted from the Orion database after the period of time designated by the Discovery Retention has passed after the time when the corresponding discovery was first defined. By default, this period is 60 days.

Network Settings

The following settings directly configure the way Orion actually polls your network.

ICMP Timeout

All ICMP (ping) requests made by the Orion poller time out if a response is not received within the period designated. By default, this period is 2500 ms.

ICMP Data

This string is included within all ICMP packets sent by Orion NPM.

SNMP Timeout

All SNMP requests made by the Orion poller time out if a response is not received within the period designated. By default, this period is 2500 ms.

SNMP Retries

If a response to an SNMP poll request made by the Orion poller is not received within the configured SNMP Timeout, the Orion poller will conduct as many retries as designated by this value. By default, this value is 2.

UCS API Timeout

All API requests made for data from Cisco UCS Manager servers time out if a response is not received within the period designated. By default, this period is 240 s.

Perform reverse DNS lookup

If you want Orion to complete revese DNS lookups for monitored DHCP nodes, check this option.

Calculations & Thresholds

The following advanced options configure the way Orion computes selected performance metrics.

Availability Calculation (advanced)

This setting designates the type of calculation Orion NPM performs to determine device availability. For more information about the two types of availability calculation: Node Status and Percent Packet Loss, see “Calculating Node Availability” in the SolarWinds Orion Network Performance Monitor Administrator Guide.

Baseline Calculation (advanced)

Upon startup, Orion NPM can calculate a baseline for the transmission rates of the various elements of your network. This baseline is used as a starting point for any comparison statistics. For more information, see “Calculating a Baseline” in the SolarWinds Orion Network Performance Monitor Administrator Guide.

Allow Secure Data On Web (advanced)

Checking this option makes it possible to see SNMP community strings in the Orion Web Console. For increased security, confirm that this setting is cleared.

Node Warning Level

Devices that do not respond to polling within the period of time designated as the Node Warning Level display within the Orion Web Console as Down. By default, this value is 120 seconds.

Counter Rollover

The Counter Rollover setting configures Orion NPM to properly handle counter rollovers. Orion NPM is capable of handling either 32‑bit or 64‑bit counters, but, by default, Orion NPM assumes counters are 32‑bit. 32‑bit counters have a maximum value of 232, or 4,294,967,296, and 64-bit counters, if they are supported by your network devices, have a maximum value of 264, or 18,446,744,073,709,551,616.

Note: The 32‑bit counters option is designated as Method 1, and Method 2 is intended fr use with 64‑bit counters. If Method 2 is selected, Orion will intentionally skip a poll if a polled value is less than the previous polled value to permit counting to 264. Orion fully supports the use of 64 bit counters; however, these 64 bit counters can exhibit erratic behavior in some implementations. If you notice peculiar results when using these counters, disable the use of 64-bit counters for the problem device and contact the device manufacturer.

I've tried to be as specific as I am without knowing exactly what you are trying to attain.

If you need more specific information, reply by letting me know what you are adjusting and why.

Hope this helps...

BSpencer

FormerMember

My initial goal is to set the monitoring values to defaults. The Solarwinds implementation I inherited was a cold hand-off in respect that I wasn't given any information from the previous admin or customers as to the how or why of the applications configuration, and I was brand new to Solarwinds product as well. So, I want to set everything back to default settings, find out what works, what doesn't work, tweaking the system until it performs satisfactorily.

From day one, the Web console has been sluggish, and many operations such as sonar discovery, time out. I've adjust the VM resources, updated the software and OS with the latest updates. The interface is markedly better than when I started , but it is still sluggish. I'm looking deeper into the application settings to see if there is anything I can change there to improve performance while polling at short enough intervals to allow for timely reports of issues.

I'm learning the application as quickly as possible, and I'm leaning heavily on videos and the community to get up to speed on all facets of the application. The sooner I am up to speed, the sooner I can write Standard Ops documents, train coworkers, and begin delegating tasks to balance the workload more evenly among the IT staff.

bspencer63

ikomrad,

Is this an inherited system or did you complete the install?

Inheriting someone else's versions of NMS is sometimes a nightmare.

All of the defaults are in my previous post, so you should be able to check and see if your polling intervals have been changed.

I would also still check to see if your polling is completing before next polling cycle as your server is "sluggish".

Also you might want to eliminate anything from being polled that is superfluous and as such not really needed. This will free up poller time and resources. Things such as Interfaces that are never used, non-critical servers, hardware that you are not interested in knowing about or reporting on anytime now or in the future. You can always re-enable polling if you find it necessary of if problems are encountered.

Videos are a great way to start out and just immersing yourself in the products helps out too as well as Thwack, as you already are finding out.

Good luck.

BSpencer

FormerMember

Thanks for all of the information and encouragement,. I agree that the videos are a great resource. I've never supported a product with instructional videos! I did inherit the app indirectly from the network team, and I was asked to clean it up and use it to consolidate all of our monitoring solutions into one tool if possible. I've tweaked settings deleted monitors and nodes that appeared to be unused.

Thanks for the resources, I've bookmarked this page as a reference during my self-training.

bspencer63

ikomrad,

Check out the SolarWinds Lab training schedule. They're very informative and a few are offered most every week.

You can also go into your customer portal, click - Support, then select Virtual Classroom. There are many videos here covering a variety of tasks within NPM and other modules. Best of all, free and you can replay as many times as you need, or pause while you catch up or do something. Good info.

BSpencer

marcher73

I suggest watching the following lab. I've watched it multiple times and reestablished my baseline upon doing so. The Geek team does a deep dive on how to fine tune many behind the scenes functionality of SolarWinds.

[VIDEO] SolarWinds Lab #38: Get Your Monitoring Squeaky Clean for 2016

pratikmehta003

Hi bspencer63

Can you throw more light in terms of the ICMP and SNMP time out value? Will this fall in between the node polling interval?

For ex: if the device is getting polled every 2 min and lets say it failed in that duration so the status goes to warning state and follows fast ping method. But then will the role of timeout start here or it would before or during the 2 min polling interval?