IP SLA-隐藏的宝石用于主动管理？

在EMA，我们花费大量时间谈论如何将齿轮转变为一种更为主动的预防方法来管理IT基础架构。无论您是查看堆栈和设备健康的下层，还是在应用程序和服务的交付方式方面堆叠，一个普遍的挑战是充分利用哪些管理数据来源可用。尽管设备性能指标的收获和NetFlow（或等效）的收集变得司空见惯，但使用IPSLA作为一组辅助测量值并不那么普遍。在本周的新闻中，Solarwinds宣布了一种新产品，Orion IP SLA经理，它将取代Orion VoIP监视器，该产品在很大程度上基于IP SLA VoIP操作集。新产品包括对所有九种IPSLA操作类型的支持，并在打开Cisco的IP SLA功能到网络管理群众的全套功能方面推动了信封。重要的是要注意，要部署Orion IP SLA Manager，您还必须购买/部署Orion Network Performance Manager。相关链接：http：//m.banksfrench.com/news/2009/092909-cisco-solarwinds.html ip SLA，IP服务级别协议的缩写，是一种嵌入式功能，是针对运行iOS的Cisco Network设备的特定功能，以及它代表的是什么是基于网络的合成测试能力。有关其工作原理以及如何设置的所有详细信息，请访问Cisco网站这里概述，以及这里有关完全更新的iOS命令参考的链接。In short, IP SLA features allow Cisco routers and switches to send various types of test messages to other IP SLA-enabled Cisco devices elsewhere in the network, so you can assess the network’s ability to deliver applications and services as well as get some idea of the quality of delivery. The nine measurement groups are VoIP, UDP Jitter, UDP Echo, ICMP Path Jitter, ICMP Path Echo, HTTP, DNS/DHCP, TCP Connect, and FTP, and there are special optimizations available for MPLS VPNs. Configuration is accomplished by CLI or SNMP, and results are harvested via FTP or SNMP. Besides SolarWinds, many other management tools have incorporated various levels of IP SLA support, and many have offered it for years. You’ll find broad support of the IP SLA operation sets within offerings from NetQoS, Entuity, Nimsoft, and Dorado. Other vendors provide partial support, such as those from CA, NetScout, Fluke, IBM, EMC, Ipswitch, SevOne, ManageEngine, ScienceLogic, PacketTrap, and (undoubtedly) others. There are even shareware and open source tools for gathering IP SLA data, such as OpenNMS and Zenoss. The most common implementation/use today is for VoIP monitoring, in particular across the WAN, from edge router to edge router. In fact, many of today’s network management tools look exclusively to the IP SLA MOS and jitter measurements to assess VoIP quality. Using IP SLA is a good preventative method for passive approximation in near real-time, and for monitoring quality when no actually traffic is flowing. This means it’s ideal for catching unanticipated impacts of new application or technology roll-outs early in their lifecycle, when corrective changes are less costly. It won’t help when you are trying to troubleshoot any individual end-user incidents, however, since it is not a measure of actual user experience. Bottom line is this – if you aren’t using IP SLA today, you should look into your existing tools to see what you can get started with. Most management tools that support IP SLA provide graphical front ends (so you can avoid the CLI configuration tasks) and orchestrated harvesting (so you don’t have to collect/collate the results yourself). If you are looking support of all nine operation sets and don’t have anything in place yet, SolarWinds has one of the few choices out there today, especially if you need to keep your getting-started costs low. Don’t be shy about checking with your existing tools vendor, however – many of them already have expanded support on their roadmaps.