Configuring/Understanding IP sla statistics - history, buckets, distributions

[SJ K]

Hi all,

I have a simple IP SLA setup using icmpecho to control / track when a route should be up or down.

It has been working fine until 1 day whereby i need to see the statistics history (due to troubleshooting purposes) and is trying to set them up.

Distribution Statistics:
   Number of statistic hours kept: 2
   Number of statistic distribution buckets kept: 1
   Statistic distribution interval (milliseconds): 20
Enhanced History:
History Statistics:
   Number of history Lives kept: 0
   Number of history Buckets kept: 15
   History Filter Type: None

I have been googling and reading the cisco command reference, but i still cannot figure out what is the actual meaning of the

a) statistics distribution ?

b) buckets ?

c) history lives ?   and the relationship between them all.

Can some kind soul / gurus lay this out in layman terms ?

Regards,

Noob

[Mark Malone]

Hi

not sure if you came across this link when searching but has a bit of information regarding your questions and found bit more i added below , havent looked that far into sla but i know history needs to be set first before it can be used

http://www.cisco.com/c/en/us/td/docs/ios-xml/ios/ipsla/command/sla-cr-book/sla_s1.html#wp2385427805

Distribution statistics--IP SLAs maintains a frequency distribution over  configurable intervals. Each time IP SLAs starts an operation, a new  history bucket is created until the number of history buckets matches  the specified size or the lifetime of the operation expires. By default,  the history for an IP SLAs operation is not collected. If history is  collected, each bucket contains one or more history entries from the  operation. History buckets do not wrap.

The distribution statistics consist of the following:

• The sum of completion times (used to calculate the mean)

• The sum of the completion times squared (used to calculate standard deviation)

• The maximum and minimum completion times

• The number of completed attempts

BucI

Bucket index number--A number uniquely identifying the statistical distribution (aggregation) bucket. The number of history buckets to be kept is configured using the historybuckets-kept command. A bucket will gather statistics for the specified interval of time (aggregation interval), after which a new statistics bucket is created. If a number-of-buckets-kept value is configured, the interval for the last bucket is infinity (until the end of the operation). Buckets are not applicable to HTTP and UDP jitter monitoring operations. This field is equivalent to the rttMonStatsCaptureDistIndex object in the Cisco RTTMON MIB.

An IP SLAs operation can collect history and capture statistics. By default, the history for an IP SLAs operation is not collected. If history is collected, each history bucket contains one or more history entries from the operation. When the operation type is ICMP path echo, an entry is created for each hop along the path that the operation takes to reach its destination. The type of entry stored in the history table is controlled by the filter-for-history command. The total number of entries stored in the history table is controlled by the combination of the samples-of-history-kept, buckets-of-history-kept, and lives-of-history-kept commands.

---

Note Collecting history increases the RAM usage. Collect history only when you think there is a problem in the network.