The device of the Stingray platform [Документация VAS Experts]

The device of the Stingray platform

Even on an unconnected server, the fastdpi_1gb process by top showed a load of around 140% (on a 4-core CPU). Is it supposed to be like that?Now at about 50Mbit, top shows 160-220% CPU Load. Is this correct or is there something else that needs to be fixed?

The high load at standby mode is due to the fact that the cards are not interrupted, but are polled continuously to ensure low latency. As data flow grows, this load is more redistributed in favor of useful work. We recommend to watch CPU load with utility mpstat -P ALL

How do I know my license expiration date?

 grep 'expiration_date=' /etc/dpi/fastdpi.lic
 expiration_date=20991231
 format: YYYYMMDD

Which files must be archived to save the license information?

/etc/dpi/fastdpi.lic
/etc/dpi/fastdpi.sig
/etc/pf_ring/*

How can I see the list of changes before installing?

On the wiki front page we publish announcements of new versions and a list of changes, or you can install the yum-plugin-changelog package and use the changelog command

yum install yum-plugin-changelog
yum changelog 4 fastdpi

Does the software only work with CentOS? We mainly use FreeBSD on our servers, is there a version for this OS?

There is no version for freebsd. Only VEОS is supported;
We strongly recommend using the OS image specified in this article.

Can I use a third-party monitoring solution?

You can use SMNP monitoring solutions such as the zabbix agent. Description

How is the load distributed across processor cores (12 cores/2 clusters)?

The cores are functionally distributed between different dpi tasks so that they do not interfere with each other.
2 and 7 are responsible for traffic transit and are currently underutilized since there is not much traffic.
You can check the distribution of functions with the command:

ps -p `pidof fastdpi` H -o %cpu,lwp,psr,comm

Why are there processor cores unloaded (the 12th core is not busy)?

The dpi has service functions: netflow generation, clickstream, pcap writing, control command processing, etc.
Their load is uneven and they can briefly load the core to 100%, so they have a separate core, so they do not interfere with the same transit.

Why can one core be 100% loaded at typical DPI load? I can't get into the server, only rebooting helps.

It is most likely the fault of kipme process, ipmi server remote control interface, maybe FW is not protecking against external attacks. When the process loads 100% kernel (any kernel) softRAID stops working, so it's not possible to log in to the server. There is also an article on other possible causes of problems with ipmi:
Kipmi0 eating up to 99.8% cpu on centos 6.4

Are the utilities url2dic and ip2bin available in source, or for FreeBSD 9 x64?

The utilities are not available in source and there are no plans to make them available in source. Freebsd allows you to run native linux application. An archive with a binary version of the utilities is also available for FreeBSD 9.2.