Examining the WCF cache using "httpalg –wcfcache"

Last modified on 15 Nov, 2022. Revision 10
Detailed output explanation of the CLI command "httpalg –wcfcache"
Up to date for
cOS Core 14.00.6
Supported since
cOS Core 9.30.x
Status OK
Author
Perter Nilsson

Question:


The output from the CLI command “httpalg –wcfcache” is a bit overwhelming, what do all these values mean exactly? I’m troubleshooting a problem related to Web Content Filtering and would like some more details about the output.

Answer:


Below is a list of all the output and explanations to what each output means:

Cache Size: 7525 URLs

The amount of URL’s in the cache. The more webpage the users are surfing to the more URL’s will be in the cache. This means that the cache will be consulted first, if it’s already in the cache the lookup will go much faster. Each URL will exist in the cache between 26 and 60hours depending on how active the particular URL is, after that it is considered outdated. If someone requests the same URL again after the URL time has expired, it will have to looked up again towards the Clavister Service Provisioning Network (CSPN), basically the Clavister update network that consists of update servers located all around the globe.

Cache Hit Rate: 3 per second.


The amount of requests that can be matched towards the local cache, instead of having to request a lookup to CSPN.

Cache Miss Rate: 0 per second.


The amount of requests that can NOT be matched towards the local cache, these requests has to be sent to CSPN. If WCF have been activated only recently the amount of misses will be initially high and later on drop once the most common URL’s are in the local cache.

Request Lookups: 0 per second.


The amount of requests we got resolved from the CSPN the last second.

Request Queue Length: 0 URLs.


The amount of URL’s in queue to be resolved towards CSPN. During this time the client request will be “stalled”.

Requests In Transit: 0 URLs.


The amount of URL’s we have sent to CSPN but not received a reply from yet. About 50-60 URL’s can be sent in each transaction.

RTT per transaction: 170 milliseconds.


Round-Trip-Time between the Firewall and the CSPN node. It can be higher/lower depending on the load of the CSPN server and the distance, it most commonly depends on the distance if a high value.

Request Queue Delta: 0 URLs per second.


This value determines how fast the URL queue increase/decrease per second. It is normal that this value would increase when WCF is activated in a busy network, it should however start to decrease after awhile and stabilize when the most common URL’s are resolved and located in the URL cache.

Cache Replacements: 0 URLs per second.


This value shows the amount of URL’s we are replacing per second if/when the URL cache is full.

Last Cache Repl. - Hit Rate Idle: N/A.


This value shows the hit rate for the latest replaced cache entry. The value has N/A in the example as the cache replacement was zero when the example command was executed.

Last Cache Repl. - Idle TTL Left: N/A.


This value shows how long before the replaced cache entry had a hit/match.

Last Cache Repl. - Session TTL Left: N/A.


This value shows how long the replaced entry would have been alive had it not been replaced just now.



Related articles

Using PCAP packet capture in cOS Core
7 Sep, 2022 core cli pcap netwall pcapdump
Avoiding cOS Core HA interruptions during configuration deployment
20 Feb, 2023 ha core idp cli cluster antivirus configuration
Automatically stop active PCAPdump or Logsnoop in the CLI
7 Dec, 2022 pcapdump log cli core logsnoop
Troubleshooting cOS Core rules/routes with ping simulation
17 Mar, 2023 core routing rules ping icmp cli



Tagscorecli