|
|
DNS Resolver Testing
Široký, Filip ; Pluskal, Jan (oponent) ; Grégr, Matěj (vedoucí práce)
This paper describes automation of creating scenarios for Deckard, which is DNS resolver testing tool. Generating scenarios build on real traffic between a web browser and a web page. The scenarios consist not only of queries from that traffic but also queries the resolver might ask for, for example, with query minimization. We should get a deterministic environment as opposed to the live environment. Live environment isn't suitable for testing due to IP address rotation, different content on servers authoritative for the same zone, content modification and so on. Also, no query should remain unanswered by a scenario. With generated scenarios, we can compare responses from different resolvers and different versions of the resolvers. This gives us a way to detect changes in behavior on a stable set of data. Also, finding or testing a bug is a possibility.
|
|
|
Dynamická konfigurace a správa modulárního DNS resolveru
MRÁZEK, Aleš
Diplomová práce se zabývá návrhem systému určeného k lokální i vzdálené administraciopen-source implementace DNS resolveru Knot Resolver, která se vyznačuje neobvyklevysokou modularitou a je realizován několika samostatnými procesy konfigurované jazykem Lua. Práce uvádí problémy spojené s administrací Knot Resolveru a navrhujeřešení v rámci systému založeného na NETCONF protokolu. Pro definici konfiguracea administračních operací je vytvořen datový model pomocí jazyka YANG. K návrhua realizaci systému je využita sada nástrojů libyang, sysrepo a netopeer2, které jsouzaložené na na otevřených standardech NETCONF protokolu a modelovacího jazykaYANG.
|
|
|
DNS Resolver Testing
Široký, Filip ; Pluskal, Jan (oponent) ; Grégr, Matěj (vedoucí práce)
This paper describes automation of creating scenarios for Deckard, which is DNS resolver testing tool. Generating scenarios build on real traffic between a web browser and a web page. The scenarios consist not only of queries from that traffic but also queries the resolver might ask for, for example, with query minimization. We should get a deterministic environment as opposed to the live environment. Live environment isn't suitable for testing due to IP address rotation, different content on servers authoritative for the same zone, content modification and so on. Also, no query should remain unanswered by a scenario. With generated scenarios, we can compare responses from different resolvers and different versions of the resolvers. This gives us a way to detect changes in behavior on a stable set of data. Also, finding or testing a bug is a possibility.
|
host ::
RSS.