More blog posts in WhiteboardWhere is this place located?
1 2 3 … 5 Previous Next

Whiteboard

65 Posts

Эндрю Хэй (Andrew Hay) недавно сделал отличную запись на веб-сайте Dark Reading, посвященную тому, является ли термин "Большие Данные" только модным словечком, используемым поставщиками SIEM с целью привлечь к себе внимание. Я должен сказать, что склонен согласиться с его выводами.  Поставщики SIEM (и управления лог-файлами), использующие архитектуры, которым уже не менее десяти лет, заявляют, что Большие Данные — это то, что они, возможно, не могут предоставить без перехода на соответствующие новые технологии и концепции.  Но мне не кажется, что эта история заканчивается здесь.

 

Концепция Больших Данных может быть использована для аналитики в любой области, начиная с финансов и заканчивая погодой. Однако большинство, скорее всего, интересуется тем, что это значит для вас, когда приходит ваш начальник и спрашивает, что вы делаете с большими данными (например, подписав документы на приобретение нового массива). Что же, давайте обсудим это.  Большие Данные применительно к ИТ могут быть использованы для управления угрозами безопасности, разрешения проблем, связанных с производительностью приложений, выявления коммерческих идей, выходящих за пределы стандартных тенденций, а также определения рабочих проблем до того, когда они станут слишком велики.  Все эти способы — достойное использование технологии и, вероятно, также вашего времени. Однако прийти к подобным идеям от Больших Данных не так-то просто.

 

Существуют две фундаментальных проблемы, связанные с переходом от Больших Данных к ценности для организации.

  1. Нужно собрать их все — как заметил Эндрю в своей записи на сайте Dark Reading, существует несколько трудностей, связанных с системами, архитектура которых относится к технологиям десятилетней давности.  Эти проблемы связаны как с системами хранения, упомянутыми Эндрю (т.е. базы данных), так и со структурой организации, используемой для хранения данных и их анализа (т.е. транзакционные схемы в сравнении со схемами оперативного анализа данных, или OLAP).
  2. Необходимо выполнять аналитическую работу, выявлять проблемы, а затем автоматизировать выявление последующих проблем. Эта задача становится еще более сложной, поскольку больших данных нет в каком-либо пошаговом руководстве, вам нужно определить их самостоятельно. Нужен специалист с обширными научными познаниями в области вычислений, чтобы найти иголку в стоге сена и определить, что она важна для вашего предприятия.  Кроме того, даже после этого определения не всегда понятно, подойдет ли используемая для анализа система для автоматизации и выявления проблем в реальном времени.  Скорее всего, это будут как минимум разные приложения.

 

Для многих из вас эти два обстоятельства превращают большие данные в непозволительную роскошь. Но если ваша задача — обезопасить сеть от угроз, или вы стараетесь предоставить высокую доступность, то вы, скорее всего, захотите продолжить чтение.   У вас есть несколько вариантов:

  • Можно сделать все сразу: Выполните сборку системы на основе таких технологий больших данных, как Hadoop и Google MapReduce, или даже на основе продуктов, больше ориентированных на платформу, например Splunk, наймите специалиста по вычислениям и пройдите весь путь от начала до конца.  Это не настолько сложно — ознакомьтесь с практическим примером
  • Можно поискать продукт, которые предоставит вам еще несколько практических средств работы.  Для начала убедитесь в том, что вы можете собирать данные компьютера и реагировать на них в реальном времени.  Если вы ведете запись данных в базу, а затем предоставляете их механизму аналитики, то это не поможет вам отреагировать достаточно быстро в случае угрозы безопасности. Затем вам нужен продукт, способный предоставить вам средства визуализации данных, такие как облака ключевых слов, древовидные структуры, пузырьковые диаграммы, гистограммы и так далее. Они помогут вам начать работу по изучению данных. Это поможет вам определить, что именно нужно искать, — сам по себе ИТ-поиск не сократит эту работу.  В-третьих, необходимо обеспечить простоту сборки правил. Пожалуйста, никакого написания вручную на языке запросов, мы живем в эпоху перетаскивания.  И, наконец, убедитесь в том, что ваша система может предпринимать действия. Если все, что она умеет, — это предупреждать вас, то она не поможет остановить проблему, она только сообщит вам о ее наличии, а это большая разница.

 

Большие Данные пришли в нашу жизнь и останутся в ней, но я советую подходить к ним с практической точки зрения, если вы не знаете, что делать в случае прекращения получения данных.  Если вы действительно знаете, что делать, то найдите решение, которые поможет вам получить львиную долю пользы без необходимости нанимать специалиста по вычислениям лично для вас.

Andrew Hay ha recentemente contribuito, con un interessante articolo su Dark Reading, alla questione che 'Grande quantità di dati' sia un'espressione diffusa dagli agenti di vendita SIEM per attrarre l'attenzione e devo dire che sono piuttosto d'accordo sull'essenza della sua conclusione.  Gli agenti di vendita SIEM (e la gestione delle sessioni) in presenza di architetture di sistemi datate di dieci anni, dichiarano che una 'Grande quantità di dati' è quella che probabilmente essi non possono consegnare fino al momento in cui non avranno familiarità con la tecnologia e i concetti delle 'Grandi quantità di dati'.  Ma non penso che con questo la questione sia terminata.

 

Mentre i concetti delle 'grandi quantità di dati' possono essere utilizzati per analizzare qualsiasi settore, dalla finanza alle previsioni del tempo, la maggior parte di voi sarà - con tutta probabilità - più interessata al significato della situazione in cui il capo vi chiede che cosa fate a proposito della 'grande quantità di dati' (probabilmente subito dopo l'ordine, da parte dell'azienda, di un nuovo array).  Per cui, parliamone.  Una 'grande quantità di dati' in campo IT, può essere utilizzata per la gestione di problemi di sicurezza, questioni prestazionali delle applicazioni, oltre che per individuare degli aspetti del business che trascendano le tendenze di base e delle questioni operative, prima che queste diventino problemi operativi.  Tutti questi punti meritano gli impieghi della tecnologia e probabilmente il vostro tempo, ma determinare questi vari aspetti a partire da una 'grande quantità di dati' non è poi così semplice.

 

Sono due i problemi fondamentali per arrivare al valore intrinseco partendo da una 'grande quantità di dati'.

  1. È necessario raccogliere tutti questi dati, come Andrew ha evidenziato nel suo articolo su Dark Reading, laddove sono svariati i problemi presenti nei sistemi la sui struttura risale ad una tecnologia datata di un decennio.  I problemi si pongono sia in termini dei sistemi di memoria, come Andrew aveva puntualizzato (vale a dire, il database), sia in termini della struttura organizzativa utilizzata per mantenere ed analizzare i dati (vale a dire, tipo transazionale nei confronti degli schemi OLAP).
  2. È necessario analizzare e individuare i problemi, per poi automatizzare tale identificazione per il futuro. Si tratta di un punto veramente difficile, dato che la "grande quantità di dati" non è accompagnata da una "guida di assistenza", ma se ne devono immaginare i vari aspetti. È richiesto qualcuno con un'effettiva esperienza nel campo della conoscenza informatica per trovare l'ago nel pagliaio e sapere che tale elemento ha un significato per il proprio business.  Inoltre, una volta individuato, non è chiaro se lo stesso sistema utilizzato per l'analisi sarà di una qualche utilità per l'automazione e il rilevamento in tempo reale.  In effetti, è probabile che si tratti di applicazioni come minimo differenti.

 

Per molti di voi, questi due fattori possono rendere una 'grande quantità di dati' un lusso scevro di aspetto pratico ma, se proprio desiderate rendere sicura la vostra rete o se volete veramente assicurare un'elevata disponibilità, forse vorrete proseguire nella lettura.   Avrete a disposizione qualche opzione:

  • Tuffarvi: costruire un sistema basato su tecnologie di 'grandi quantità di dati', quali Hadoop e Google MapReduce, o addirittura un numero ancora maggiore di prodotti orientati alla piattaforma, quali Splunk, assumere il vostro scienziato informatico e proseguire nel cammino.  Non è impossibile: date un'occhiata a quest'analisi di studio
  • In alternativa, potrete ricercare un prodotto che vi possa fornire un certo numero di altri strumenti pratici.  Per prima cosa, assicuratevi di poter raccogliere dati sulle macchine e di agire su tali dati in tempo reale.  Se state trascrivendo i dati verso il database, rendendoli quindi disponibili ad un motore di analisi, questo non vi aiuterà a reagire con sufficiente rapidità ad una minaccia nel campo della sicurezza. Secondo punto: ottenete un prodotto che possa offrirvi alcuni strumenti di visualizzazione dei dati, fattori quali "cloud", mappature ad albero, diagrammi a bolle, istogrammi e via di seguito vi aiuteranno ad avviare un esercizio di esplorazione dei dati. Questo vi aiuterà a definire gli obbiettivi della ricerca: la ricerca IT di per sé non eliminerà questa necessità. Terzo punto: assicuratevi che la definizione di regole sia facile; nessun linguaggio di interrogazione redatto a mano, per favore, dato che siamo nell'epoca del trascina-e-rilascia.  Ultima fase: assicuratevi che il vostro sistema possa intraprendere azioni, se tutto quello che può fare non è veramente di aiutarvi ad arrestare il problema, ma semplicemente di aiutarvi a sapere che sussiste un problema: si tratta di una grande differenza.

 

La 'grande quantità di dati' è presente per restare, ma il mio consiglio è di essere pratici, se non sapete quello che volete fare in caso di arresto dei dati.  Ma, se sapete bene quello che fareste in tal caso, trovate allora una soluzione che vi procuri le parti chiave del valore, senza dover assumere il vostro scienziato informatico particolare.

Andrew Hay a rédigé un post intéressant sur Dark Reading récemment. Il se demandait si les données massives n'étaient véritablement qu'un buzz initié par les fournisseurs de SIEM pour attirer l'attention. J'aurais tendance à me poser la même question. Les fournisseurs de SIEM (et de systèmes de gestion des journaux), qui travaillent sur des architectures qui ont plus de 10 ans et se targuent de savoir gérer les données massives, ne sont certainement pas en mesure de le faire à moins d'avoir adopté les nouveaux concepts et les nouvelles technologies de gestion de « Big Data ».  Et, à mon avis, l'histoire ne fait que commencer.

 

Bien que le concept de « big data » soit un peu fourre-tout (puisqu'il permet d'analyser des données financières comme des données météorologiques), la plupart d'entre vous veulent surtout savoir ce qui se cache derrière ce concept, en particulier lorsque votre patron vous demande comment vous gérez les données volumineuses (juste après avoir signé un contrat portant sur la fourniture de l'un de ces systèmes, en général !). Alors, parlons-en ! Les « big data » peuvent : aider à réagir à des menaces informatiques ou à des problèmes de performance d'une application, permettre d'identifier des pratiques qui diffèrent des tendances du moment et d'identifier des problèmes opérationnels avant qu'ils ne se posent réellement. Tout ceci justifie donc des investissements en termes de technologie, et de temps, mais obtenir des informations exploitables à partir des données massives n'est pas une mince affaire.

 

Deux problèmes majeurs se posent :

  1. Premièrement, vous devez collationner toutes les données, et Andrew souligne à juste titre dans son post sur Dark Reading, que des systèmes qui s'articulent autour d'une technologie vieille de dix ans ne sont pas sans poser des problèmes, tant en termes de stockage (base de données) que de la structure organisationnelle employée pour conserver et analyser les données (schéma transactionnel contre schéma OLAP).
  2. Il vous faut donc analyser le concept, identifier les problèmes et automatiser l'identification pour la fois suivante. Ceci ne va pas sans difficulté car les données massives ne sont pas fournies avec un guide ! Vous devez trouver la solution tout seuls. Vous devez donc disposer d'une réelle expertise en calcul scientifique pour trouver l'aiguille dans la botte de foin et savoir que tout ceci fait sens pour votre activité. D'autre part, une fois le concept identifié, rien ne dit que le système employé pour l'analyse donnera de bons résultats en termes d'automatisation et de détection en temps réel. En réalité, il s'agira au moins d'applications différentes.

 

Pour un grand nombre d'entre vous, ce début d'explication rend les big data difficilement accessibles à votre niveau mais, si vous travaillez à la sécurisation de votre réseau ou à assurer une réelle disponibilité, lisez la suite de l'article. Plusieurs options s'offrent à vous :

  • Sautez dans le bain à pieds joints : Créez un système basé sur des technologies de données massives (Hadoop ou Google MapReduce, ou des produits encore plus orientés plate-forme comme Splunk par exemple), embauchez un expert en calcul scientifique et lancez-vous. La tâche sera difficile mais pas impossible. Pour avoir une idée, lisez l'étude de cas
  • Vous pouvez aussi chercher un produit qui offre des outils plus simples à utiliser. Pour commencer, vous pouvez collecter des données machine et les traiter en temps réel. Si vous écrivez les données dans la base de données, en les rendant disponibles pour un moteur d'analyse, vous ne pourrez pas réagir assez vite à une menace informatique. Deuxièmement, en choisissant un produit doté d'outils de visualisation des données (clouds, Treemaps, graphiques à bulles, histogrammes, etc.), vous pourrez mieux aborder votre exercice d'exploration des données. Vous saurez ainsi mieux quoi chercher (une simple recherche informatique ne suffira pas à obtenir des résultats probants). Troisièmement, vérifiez qu'il est possible de créer des règles simplement (par pitié, pas de langage d'interrogation manuscrit, nous sommes à l'âge du glisser-déposer). Enfin, vérifiez que votre système peut prendre des mesures (s'il ne peut que vous alerter, il ne vous aidera pas à solutionner le problème mais juste à vous signaler qu'il y a un problème, et cela fait une énorme différence.

 

La gestion des big data est là pour un moment. Le seul conseil que je peux vous donner est d'adopter une approche pratique et de trouver une solution qui vous soit profitable sans devoir embaucher un expert en calcul scientifique.

In einem großartigen Artikel auf Dark Reading fragte Andrew Hay unlängst, ob Big Data nur ein Modewort ist, das SIEM-Verkäufer nutzen, um auf ihre Produkte aufmerksam zu machen, und ich muss sagen, dass ich ihm nur zustimmen kann.  Verkäufer in den Sparten SIEM (und Logmanagement) mit 10 Jahre alten Architekturen, die behaupten, dass diese mit Big Data umgehen können, können ihr Versprechen nicht einhalten, außer sie wenden neue Technologien und Konzepte für Big Data an.  Aber ich glaube, das ist noch gar nicht alles.

 

Während mithilfe von Big Data von Finanzdaten bis hin zum Wetter alles analysiert werden kann, interessiert es Sie wahrscheinlich eher, was es für Sie bedeutet, wenn Sie Ihr Vorgesetzter fragt, was Sie wegen Big Data zu unternehmen gedenken (wahrscheinlich direkt, nachdem er den Kauf eines neuen Arrays genehmigt hat).  Reden wir also darüber.  Im IT-Bereich kann mithilfe von Big Data vieles gemacht werden: Sicherheitsbedrohungen und Probleme bei der Anwendungsleistung verwalten, Geschäftseinblicke über grundlegende Trends hinaus erhalten und operative Schwachstellen erkennen, bevor sie operative Probleme werden.  So lässt sich nicht nur die Technologie, sondern auch Ihre Zeit durchwegs gut nutzen. Aber es ist nicht so einfach, Einblicke aus Big Data zu erlangen.

 

Bei der sinnvollen Nutzung von Big Data gibt es zwei fundamentale Probleme:

  1. Zum einen müssen Sie alles erfassen – wie Andrew in seinem Artikel auf Dark Reading sagt, gibt es mehrere Probleme mit Systemen, deren Architektur auf einer zehn Jahre alten Technologie basiert.  Die Probleme beziehen sich auf die Speichersysteme, die Andrew nennt (d. h. Datenbank), sowie auf die Organisationsstruktur, um die Daten zu führen und zu analysieren (d. h. Transaktions- ggü. OLAP-Systeme).
  2. Zum anderen müssen Sie Probleme analysieren und identifizieren und anschließend die Identifizierung hinkünftig automatisieren. Das ist noch schwieriger, denn für Big Data gibt es keine Anleitung. Sie müssen schon selbst dahinterkommen. Dazu braucht es einen ausgebildeten IT-Spezialisten, um gerade jene Nadel im Heuhaufen zu finden, die für Ihr Unternehmen wichtig ist.  Und wenn sie einmal gefunden wurde, heißt das noch nicht, dass das für die Analyse verwendete System auch für die Automatisierung und Realtime-Erkennung zu gebrauchen ist.  Die Wahrscheinlichkeit ist sogar hoch, dass es sich dabei zumindest um verschiedene Anwendungen handelt.

 

Aus diesen zwei Gründen sind Big Data für viele ein unpraktischer Luxus. Sollten Sie jedoch Ihr Netzwerk sichern oder hohe Verfügbarkeit erreichen wollen, dann lesen Sie bitte weiter. Sie haben folgende Möglichkeiten:

  • Packen Sie's an: Bauen Sie ein System auf Basis von Big-Data-Technologien wie Hadoop und Google MapReduce oder auch mehr plattformorientierter Produkte wie Splunk, stellen Sie einen IT-Spezialisten ein und legen Sie los.  Das ist gar nicht so schwer – wie diese Fallstudie
  • Oder Sie sehen sich nach einem Produkt um, das Ihnen ein paar praktischere Tools an die Hand gibt.  Erstens: Achten Sie dabei darauf, dass Sie damit Maschinendaten erfassen und in Echtzeit darauf reagieren können.  Wenn Sie die Daten in eine Datenbank schreiben und Sie erst dann einer Analysemaschine zugänglich machen, können Sie nicht schnell genug auf eine Sicherheitsbedrohung reagieren. Zweitens: Besorgen Sie sich ein Produkt mit Visualisierungstools für die Daten, z. B. Wortwolken, Treemaps, Bubblecharts, Histogramme usw., welche die Untersuchung der Daten erleichtern. So erkennen Sie leichter, wonach Sie suchen sollten – die IT-Suche alleine schafft das nicht.  Drittens: Die Erstellung von Regeln muss einfach sein – bitte keine handschriftlichen Abfragesprachen, wir sind im Zeitalter von Drag-and-Drop.  Schließlich muss das System in der Lage sein, Aktionen durchzuführen. Wenn es Sie nur warnt, hilft es Ihnen bei der Problembehebung nicht. Es zeigt Ihnen nur, dass es ein Problem gibt – das ist ein großer Unterschied.

 

  An Big Data führt kein Weg mehr vorbei. Mein Rat lautet: Gehen Sie die Sache praktisch an. Wenn Sie nicht wissen, was Sie mit den Daten tun sollen, lassen Sie's.  Wenn Sie wissen, was Sie damit tun sollen, finden Sie eine Lösung, deren Funktionen es auch wert sind - damit Sie keinen eigenen IT-Spezialisten anstellen müssen

The amount of noise in the virtualization and private cloud marketplace right now is absolutely deafening – even when you look at small segments of the market. There are now at least four hypervisors in relatively broad use. There are at least four different companies offering backup solutions for virtualization. Every storage vendor on the planet is developing one (if not all) of their products specifically to support a virtual infrastructure. Let’s not even talk about the number of companies, both large and small, that are developing products for the “Cloud.”

 

One of the most confusing areas in the market right now is virtualization management. Off the top of my head, I can think of at least six companies that advertise a virtualization management tool. There are several different layers of this solution worth considering.

 

We all know about the hypervisor – the tool that makes true abstraction from the physical layer possible. VMware says they give you the hypervisor (ESX) for “free.” Microsoft includes Hyper-V with the Windows Server OS, and KVM & Xen are both open source (free) hypervisors.

 

Next, there’s the orchestration layer. This layer manifests itself differently depending on which hypervisor you’re managing. VMware used to be the only game in town, but there are several folks pushing into this market now. The open source OpenStack project founded by Rackspace and NASA a couple of years ago shows the most promise as an orchestration layer competitor to VMware, but there are several smaller players in the market that could get some traction.

VMware charges you for layered functionality on top of ESX by requiring vCenter to unlock the real benefits of virtualization – VMotion, DRS, Storage DRS, HA, etc. Many people view vCenter as a virtualization management solution, and don’t even know that there is an entire world of tools out there that complement vCenter to give you some really valuable features that vCenter does not provide. Even worse, I think there are still a lot of folks out there who think they have to stick with a 100% VMware virtualization stack in order to be “enterprise class.”

 

Microsoft has taken a different angle. There isn’t a vCenter equivalent for Microsoft’s Hyper-V hypervisor. Microsoft gives you much of the same functionality for free that VMware requires vCenter to utilize. It’s true that, today, Hyper-V is missing some of the great capabilities of VMware, but taking just a short leap to the end of 2012, when Hyper-V 3.0 launches, a new chapter will begin. Hyper-V 3.0 promises to be a game-changer that could make the top virtualization platforms interchangeable from the standpoint of functionality.

 

However, there are a bunch of new tools out there that provide metrics, functionality, and intelligence that you just can’t get from the tools most environments employ from Microsoft and VMware. These virtualization management solutions help you with the heavy lifting in managing VMware and Microsoft environments by providing you with real-time dashboards, reports, alerting, tools for capacity planning and analysis, performance monitoring, bottleneck detection, and so much more. There are several reasons all of these third-party players are developing virtualization management tools:

  1. VMware has chosen to build a VMware-only monitoring solution in vCenter Operations Manager. This equates to VMware proprietary lock-in from the only folks really capable of pulling it off. Don’t be fooled!
  2. vCenter alone is just orchestration, and many virtualization administrators want better metrics, analytics, reporting, and other functionality than it can offer.
  3. Microsoft Systems Center Virtual Machine Manager (SCVMM) is mainly focused on provisioning. Again, there are lots of other functions necessary for better virtualization management.
  4. BOTH of the solutions above are very expensive…especially when you want the full functionality that many of the smaller players offer. This opens the door to a ton of value solutions like SolarWinds Virtualization Manager that are priced at a fraction of competitive tools.

 

SolarWinds will even let you test drive Virtualization Manager for free with a trial for an unlimited number of virtual machines for 30 days. Then, when you’ve decided Virtualization Manager is the best platform for your environment, you’ll be surprised to find out that it costs substantially less than most competitive tools. Substantially! This is one of the few situations where you can have your cake and eat it, too!

In a recently published article, “Forget Improvements, Systems Management Needs a New Approach,” Denny LeCompte and I argued that the challenge around adopting application performance management is largely organizational.  Software companies can help remove these organizational barriers with easy to purchase, deploy and use software.  One of our competitors, NetIQ. agrees with our assessment that usability issues hold back success in IT management.  In a recent NetIQ blog, written by Travis Green, he indicated usability was the focus of their latest release of AppManager v8, although to back up his claim, he describes a somewhat obscure, “big enterprise” use case.

 

Perhaps NetIQ uses the term differently, but when we refer to usability, we mean that the product can be learned quickly and easily, that the daily use of the product is efficient, and the experience of the product is satisfying and pleasing.  Inextricable from usability is whether the product provides the proper features to solve the problems for which it is intended.   SolarWinds is maniacal about our focus on getting the features right.  We don’t care about winning irrelevant battles between software feature checklists.   We simply won’t add a new feature because one big customer wants it; we only add features that hundreds of our customers need.  In the end, SolarWinds uses the acid test of usability:  If our products were not easy to use, SolarWinds would not make any money because every prospective customer downloads, installs, and deploys the software all on their own.  The truth is that we don’t have any professional services staff to do it for them.

 

Getting a product to be usable requires focused effort and an user-centric approach to development.  What this means is providing frequent and varied means of customer feedback and interaction such as:

 

·         Usability Tests

·         Iterative, user-centered design

·         Customer experience interviews

·         Beta programs with high participation

·         An open forum (like thwack) for customers to criticize, praise, or explain their needs

·         and frequent product releases.

 

Over the last four years, SolarWinds Server & Application monitor (SAM) has iterated on usability improvements using all of these methods, providing one major release and one minor release each year.   In fact, I spoke with one customer this morning and he told me that the latest version of SAM (which shipped in March, 2012) is now one of the best products on the SolarWinds Orion base.  What an accomplishment for a product with such a short life span!

 

Take the Technology Taste Test

I am curious to hear how NetIQ AppManager v8 customers like the new release.  I am also curious when NetIQ will iterate on these improvements.  For the sake of their customers, I hope the time span will not be as great as in previous releases (AppManager v7 GAed in March 2007, or nearly 5 years prior to v8).  Better yet, NetIQ customers, Enterprise Systems Journal readers or anyone else should compare SolarWinds’ usability against NetIQ.  Download both products and  I have a pretty good idea who will win the blue ribbon, but, like I said, SolarWinds is open to constructive criticism.

Those famous words from one Rowan Atkinson from the Blackadder series (if you don’t know Blackadder, please go to youtube) have been front of mind in recent days.  The reason?  Well, as we compete for your business there are inevitably requests to compare our products to those of our competition.  Whether it’s Splunk vs Log & Event Manager or InfoBlox vs IPAM, or VMware vCenter Ops vs Virtualization Manager you as IT consumers always want to know what’s the best choice for your situation.

 

Of course as marketers ( AKA ministers of truth J), we are always looking for the angle that communicates why we believe our product to be the best for you, but sometimes we forget that we should let the products do the talking.  One thing that I love about the way we market and sell at SolarWinds is that ultimately the products actually do “speak” for themselves.  It’s a model we’ve all gotten used to over here, but for those of you who buy IT products I forget that the model is still somewhat unique. How often do you get to take a product you’re evaluating, put it in your business – or your home for that matter – and keep it there for 30 days, to decide if it works for you?  Of course in those 30 days we will ask you to buy it because we hope you like it, but if you wanted the full 30 days to think about it, the time is yours to use.

 

Some folks call what we do freemium (since we have free tools), but it’s not really.  Our free tools are completely separate and disconnected products from those we sell.  There are companies that use the freemium model, like Splunk, who give you a product free up to a point and then charge you for it.  But our belief is that free is free, and if you have a paid product then we’re going to let you try it, but ask you to pay for the value you receive after you’ve tried it and like it.

 

So, it’s back to my competitive plotting, but there you have my cunning plan – it’s not so cunning after all – see a product, try a product, and then buy a product – that is how I like to win. 

In Matthew Jones' blog post last week, he presents a good overview of the challenges that today's organizations deal with in regards to patching 3rd party updates when Microsoft Windows Server Update Services (WSUS) is the chosen patching mechanism. The most significant challenge being the simple fact that they're not getting patched, or if they are, it is likely in a haphazard manner at the whim of an end-user sufficiently motivated by desktop popups from the auto-updaters for those products.

 

While the article does a great job of calling attention to the problem, and even offers some suggestions for improving the environment, it doesn’t really provide a functional solution to the reader. For example, regarding educating users, the whole idea of a centralized patch management product is that users don’t have to be ‘educated’ -- an effective patch management system is completely transparent to the end user. Avoiding the expectation that users will install updates is exactly the reason the organization has implemented WSUS in the first place.

 

In the last paragraph, Jones offers the recommendation to "...implement a patch management solution that will deploy third-party patches", and provides two options, only one of which can actually be used in a WSUS environment. Other options do exist, but seem to have been overlooked in the article.

 

For the reader who is managing a WSUS environment, one product certainly worthy of mention is SolarWinds Patch Manager.  Patch Manager sits on top of the WSUS environment, provides automatic synchronization to a catalog of ready-to-use third-party updates for all of the prevalent desktop applications: Adobe Reader, Adobe Flash, Firefox, Chrome, Java Runtime, and iTunes, to name a few. In addition, Patch Manager provides an enhanced toolset for monitoring and managing the entire WSUS environment, and a toolset to directly deploy on-demand, or explicitly scheduled, third-party updates and Microsoft updates. Patch Manager also provides tools for asset inventory and reporting on the actual state of the products and updates in the organization, and it does all of this at a price point less than a third of the other option noted.

 

Patching third-party content should be no different at all from patching Microsoft content. The only reason it would be is because the methodologies are different (e.g. using GPO/Software Distribution, or trusting users to click on the auto-updater). With WSUS, the policies should be identical. More so, with WSUS, you don’t have to “scan” systems to get information – it happens automatically, daily. Publish the third-party update to the WSUS server after it automatically arrives, and in the morning run a report (or schedule it and deliver it via email to your Inbox) and review the status of your third-party updates side-by-side with your Microsoft updates.

Over the last few weeks those of you who are members of the SolarWinds community, thwack, have seen a number of changes that culminated with the switch over to the new platform that will carry our community efforts forward.  The technology changes are only the start of where we’re headed and we hope to make community a much bigger part of your relationship with us, with a whole set of new content and tools.  But before I go there, let’s talk about the vision we have for thwack.

 

For a number of years now, thousands of you have interacted with the SolarWinds product managers and support team to get the most out of our products.  You’ve given us a ton of valuable feedback, celebrated our successes with great releases, and beaten us up if we haven’t lived up to your expectations. Frankly, you’ve been integral to our ability to build great products.  But for some time, we have wanted to expand the dialogue beyond our products into more thought provoking content, educational content, useful tools, etc.  You know, look beyond our backyard…

 

Some of this has already started, for those of you who don’t know, we now have 3 blogs at SolarWinds:

 

  • Product blog: You guessed it; we talk about products here, what’s new, how to do interesting new things – in general how to get the most out of the product.  If you’re an NPM or SAM (formerly APM) customer the product blog feed is available directly in the product.
  • Geek Speak: This is the place to learn about all the ins and outs of technology, we try to make sure the content here is all about the technology and technology events, tips and tricks… really the nuts and bolts content.
  • Whiteboard: Our corporate blog (it’s not as stuffy as a corporate blog might sound) – this is our perspective on the IT market the winners, the losers, who’s dressed well on the red carpet etc. (But, you have probably already figured this one out...  since you are here, reading this post.)

 

But the blogs are just the start of what we see ahead, we’ve got exciting free tools in the works, posts from SolarWinds ambassadors (folks outside of SolarWinds), new incentives to get you to explore some of this stuff.  And that is only the beginning; the community team has some other “fan-favorite” ideas they are putting into motion. The idea is to make thwack a place where you can come to learn and interact with other IT users – it’s bigger than just SolarWinds and our products, it’s about IT and it’s about you.

 

So we’re at the beginning, but I encourage you to take a look around, tell us what you like, tell us what you want to see, and know that we’re working to take community to a whole new level at SolarWinds so if you haven’t jumped in, now’s the time.

If you read the SolarWinds blog and have seen my recent posts, it won’t be any surprise to you that we believe the virtualization market is going through a period of transformation. We believe this because our customers are telling us that, while VMware is still a significant portion of their environment, they’re beginning to consider alternatives. We could (and have) spent days hypothesizing about the reasons for this shift, but I’d rather just talk about what SolarWinds is doing to accommodate it.

 

I’m excited to announce that SolarWinds Virtualization Manager now supports Microsoft Hyper-V! There will still be a few folks out there who wonder why we decided to do this, so here is a little bit of our thought process:

  • Microsoft Hyper-V adoption is growing faster than any other hypervisor in the market right now. This one is pretty obvious.
  • We believe there is a contingent of organizations out there who remember, or are still experiencing the many pains of vendor lock-in and want to have a dual hypervisor strategy in order to avoid it going forward.
  • There are companies that don’t trust their infrastructure vendor to tell them how much new infrastructure they need to buy. Many think that VMware vCenter Operations telling them that they need to buy more VMware vSphere licensing is kind of like having a fox guarding the hen house. So, there is a viable market for a third-party, unbiased virtualization management tool that doesn’t have a horse in this race.
  • There is a huge base of Microsoft administrators out there that will help accelerate adoption.
  • Microsoft’s planned enhancements proposed for Hyper-V version 3 this year have more and more customers planning for a dual hypervisor environment.
  • Multiple management consoles are hard to use, clunky, and don’t offer you the flexibility you need in measuring and managing your virtual infrastructure. Virtualization Manager now allows you to see both VMware and Hyper-V environments in a single pane of glass.

 

Since we already know that Hyper-V penetration is accelerating, SolarWinds has a unique ability to bridge a major Hyper-V functionality gap in storage visibility. Because we’ve created integration capabilities for SolarWinds Storage Manager within SolarWinds Virtualization Manager, you can now see what was previously invisible in your Hyper-V environment. When using these two products together, you can drill down from the Hyper-V (or VMware) VM to the storage LUN servicing that VM, you can see all of the other VMs serviced by that LUN (to help you pinpoint if a particular VM is hogging storage I/O), and you can even see the physical disks associated to the LUN.

 

So, if you think any of these points are valid, you should give the new SolarWinds Virtualization Manager 5.0 a try today with our free 30-day trial today!

Andrew Hay recently did a great post on Dark Reading on whether big data was really just a buzzword being thrown around by the SIEM vendors to get attention, and I have to say I tend to agree with the essence of his conclusion.  SIEM (and log management) vendors who have 10 year old architectures claiming big data is what they do probably can’t deliver unless they’ve embraced new big data technology and concepts.  But, I don’t think that’s the end of the story.

 

While the concepts of Big Data can be used to analyze everything from finance to the weather, most of you are probably more interested in what it means to you when your boss asks you what you’re doing about big data (probably right after they sign off on buying that new array).  So let’s talk about it.  Big Data for IT can be used to manage security threats, application performance issues, identify business insights that go beyond basic trends, and identify operational issues before they become operational problems.  All of these are worthy uses of technology and probably your time but getting these insights from big data ain’t so simple (“ain’t” is the little bit of Texas coming out in me).

 

There are 2 fundamental problems in getting to value from big data.

  1. You need to collect it all – as Andrew pointed out in his Dark Reading post there are several issues with systems architected around decade old technology.  The problems are both in terms of the storage systems which Andrew pointed out (i.e. database) and the organizational structure used to keep and analyze the data (i.e. transactional vs OLAP schemes).
  2. You need to analyze, identify problems, and then automate the identification for the future. This is more challenging because big data doesn’t come with a “walk through guide”, you have to figure it out. It requires someone with real computational science expertise to find the needle in the haystack and know that it’s meaningful to your business.  Also, once identified it’s not clear that the same system used for analysis would be any good for automation and real-time detection.  In fact it’s likely that they are different applications at a minimum.

 

For many of you these 2 things may make big data an impractical luxury, but if you’re in the business of securing your network or really trying to deliver high availability you may want to read on.   You have a few options:

  • Jump in with both feet: Build a system based on big data technologies like Hadoop, and Google MapReduce, or even more platform oriented products like Splunk, hire your computational scientist and go at it.  It’s not impossibly hard – check out this case study about Zion’s Bank.  But set your expectations right because this isn’t going to be a deploy it, gather data, and voila results kind of project.
  • Or you could look for a product that can give you a few more practical tools.  First make sure you can collect machine data and act on it in real-time.  If you are writing the data to the database, then making it available to an analysis engine it’s not going to help you react fast enough to a security threat. Second, get a product that can give you some visualization tools for the data, things like word clouds, Tree maps, bubble charts, histograms etc will help you begin your data exploration exercise. This will help you figure out what to search for – IT search by itself isn’t going to cut it.  Third, make sure it’s easy to build rules – no handwritten query languages please, we are in the age of drag-and-drop.  Lastly, make sure your system can take actions, if all it can do is alert you then it’s not really helping to stop the problem it’s just helping you know that there’s a problem – big difference.

 

Big Data is here to stay, but my advice is to get practical, if you don’t know what you want to do with the data stop.  If you do know what you want to do with it then find a solution that gets you the key parts of the value without having to hire your very own computational scientist.

So I’ve been reading a lot of commentary on the impact of BYOD and even more opinions on what should be done, stop it, help it, control it, manage it, ignore it.  In fact, we recently worked with Network World on a survey around this topic.  While I understand the perspectives on what to do, in all that I’ve read I have yet to see someone approach this from the customers’ point of view.  I have been thinking about what the reason is for this movement and I’ve bounced around a few places.

 

  1. It’s just a continuation of the consumerization of IT – we saw it with the adoption of SaaS and this is just a sequel.  Users saw that they could push new technology on the business without the cooperation or consent of IT.
  2. It’s cool – let’s face it, many of these devices started out as a new toy and now we just want to use the new toy at work.
  3. It’s about easy, like the Staples Easy button, knowledge workers are busy folks and have a million things going on and any way to help make life easier is welcome.  It turns out my tablet made my personal life easier so why don’t I see if it can make my corporate life easier.

 

I think the real answer is that there’s a little truth to each of these reasons.  I’m going to start with the last one first because I think it’s the most relevant.  Work has blended into our personal lives in ways that didn’t exist 30 years ago.  The generations in the workplace today walk around never unplugged from the office.  So after pushing us to stay connected it’s not surprising that when I find an easy way to solve a data management (it’s all data) problem in my personal life I want to bring it to work.  And saying no is not an option for IT, mostly because the biggest demand for a solution is coming from the executives.  A recent survey we conducted at SolarWinds illustrates this.  When asked about the results of allowing personal mobile devices on the network 51% of respondents said it increased productivity and about 50% said it increased the ability to work from home.

If you believe my premise that the users have found a better way, then the question becomes what do we as an IT organization need to do to enable the users while still hitting on all the rules and requirements that we’re required to enforce (compliance, security etc).  So far, for IT organizations that have allowed personal devices on the network has meant more traffic (40% said this in our survey) and more helpdesk requests (44%).  So let’s peel the onion back a bit on what might make life a bit easier.

 

The first thing to realize is that users don’t need everything (they might want it though).  What users need is access to their primary sets of data – what do I mean by this?  Well not too many folks want to open massive spreadsheets, build ppt decks from scratch, or do compute intensive tasks on their tablet – there are a few, and your requirements may vary, but I’m going to propose that there’s a good 80% of users who are knowledge workers that don’t want to do the things I mentioned above. What they do want is access to email, key corporate apps (many of which are already accessible through a browser with a VPN), and maybe a few special data sets (operational reports).  So if we scale the requirements back to this then can you build a better mouse trap?  I think so.

 

I won’t go through all the options in this post because most of them aren’t new, but if you’re thinking about BYOD before you think about the answer, see if you can put a finger on the problem.As a side note, here are a couple of good pieces I read on BYOD that talk about some of the challenges and the ‘what’.

 

The new iPad has CIOs quaking in their cubicles – GigaOm

Bring your own device debate – ZDNet

3 Predictions on the Future of Enterprise Software -TechCrunch

We’ve all heard the story: organizations are standardized on VMware, and they’re not going to change. IT departments live and die by the scale they build into their environment. They daily suffer the consequences of previous regimes that left them managing a complex ecosystem of hardware, OSes, applications, etc. Based on that, we KNOW that no organization will want to add the complexity of multiple hypervisors to their virtual environments…OR DO WE???

 

A Q3 2011 survey by independent market research company Vanson Bourne indicated that 38 percent of businesses are planning to switch hypervisor vendors within the next year due to licensing models and the robustness of competing hypervisors. We’d be naïve to think that that means organizations are totally removing VMware from their environment. The fact is that VMware just works, and it works very well. That’s why they are almost always at the center of these discussions. They have owned the market, and are the incumbent in most situations. We’re really pretty comfortable with, and maybe even accustomed to, deploying mission critical applications on VMware vSphere virtual infrastructure. However, there is a growing noise in the virtualization marketplace. Companies, large and small, are talking about Microsoft Hyper-V. But, why? I believe one major reason is because of the hype surrounding Hyper-V 3.0.

 

But first, let’s step back for a second and look at why companies are looking at Hyper-V, even today, presumably months before the 3.0 release. I see two primary drivers for this consideration:

 

  1. Organizations realize that VMware is getting a lock on their business, and that monopoly power makes people nervous – they want to keep a second option open to improve their negotiating position for their next contract with VMware or in case VMware continues to raise their prices even more.
  2. Companies want a hypervisor vendor and a roadmap that will not make them regret their investment – RedHat or OpenStack don’t give them that confidence yet because they’re relative unknowns in the virtualization landscape. Only VMware and Microsoft have an established track record here.

 

To date, Hyper-V has not been able to compete head-to-head with VMware from a technical standpoint. VMware had a big jump on Microsoft, but Microsoft has been investing heavily in the last few years to catch up. While Microsoft hasn’t bridged the technical gap yet, there is lots of buzz about upcoming functionality in Hyper-V 3.0 that will add key functionality that will help them cross the technical chasm. The major features expected in Hyper-V 3.0, expected out later this year, are as follows:

 

  1. Improved Scalability – larger cluster support, native NIC teaming, and new VHDX file that will support up to 16 TB virtual disks. This enables Hyper-V to handle more mission-critical applications and workloads – particularly for Microsoft applications like, like SQL & Exchange.
  2. Extensible Virtual Switch – providing partners the capability to build in advanced networking features, like capture extensions lets them build tools to monitor and sample traffic. This also creates opportunities for networking vendors to create third party virtual switches like the Cisco Nexus 1000V for VMware.
  3. Live Migration Enhancements – Live Migration will perform much more like vMotion with live storage migration (previously, this required downtime) and concurrent live migrations of virtual machines. Microsoft may gain an advantage on VMware in this area because shared storage is not required for Live Migration. This could allow Hyper-V access to the environments with no shared storage such as small businesses and large public clouds. VMware and Xen require shared storage. RedHat’s RHEV is the only other hypervisor that does not require shared storage for live migration.
  4. Replication - embedded host-based replication will help Microsoft make Hyper-V a better fit for mission-critical applications and enables organizations to use branch offices as failover targets.

 

So, it’s obvious that Microsoft is extremely focused on competing with VMware technically. Hyper-V will still needs some capabilities strengthened to be a viable competitor to VMware’s DRS and other advanced functionality in more sophisticated environments. However, most moderately-sized and smaller organizations aren’t ready to actually implement this technology yet. So, Hyper-V still has some time to catch up. We’ll just have to keep an eye on their roadmap to see what features they’re planning to introduce next!

 

In the meantime, we’ve just released a new free tool called VM Monitor for Hyper-V that will allow you to start gaining some visibility into a Hyper-V server (we also have a VMware version of the same tool called, you guessed it, VM Monitor for VMware). Stay tuned next week for more exciting announcements around support for Hyper-V in SolarWinds products (wink, wink)!

A popular trope in science fiction involves an individual who’s been time-shifted into a future culture (frozen in ice, cryogenic sleep, time machine, what have you), and the reader or viewer then learns about the new culture by what the protagonist experiences or reads.  I’ve been thinking that if our time traveler arrived were a sysadmin who arrived in 2012 from a 5 or 10 years in the past, he or she would initially read the tech blogs and online press and think that cloud was the reigning technology and that servers in data centers and with actually applications running on them disappeared like so many 8-track tape players.

 

Of course, if the story were allowed to continue, at some point our time-traveler would get a job, maybe as a sysadmin, and then shock would set in when the cloud-enabled world was hardly to be found.  The server room would still be full of racks of physical servers.  Sure, lots of them would be running a hypervisor with virtual machines, but it’s still a far cry from the cloud-filled IT world that the hype seems to imply.

 

Am I saying the cloud is all hype?  Definitely not.  But like most technologies, there’s what some have termed macro-myopia, which is the tendency to overestimate the short-term impact of a technology and dramatically underestimate its long-term impact.  It’s clear that cloud will impact IT in a huge and probably deeply transformational way.  Eventually.   But today, most computing resources are still on solid ground.

 

How do we know?  First off, SolarWinds sells enterprise software at a really low price, which results in literally thousands of transactions every quarter.  On top of that, our product management team is reaching out to customers and non-customers alike on a daily basis, asking for what they need.  In all of these conversations, we’re focused on current pain points.  We discuss the problems that need immediate attention.  While we now have a large portfolio that can solve a lot of pains, we still get asked for things we don’t do (yet).  So how often do we get asked for help with managing cloud infrastructure?  I won’t say never, but if I did, it wouldn’t be far off.

 

We’ve also done more formal data collection.  When we surveyed 90 customers, we found that 56% of customers said they weren’t running anything in public or private clouds.  About 29% said they aren’t even thinking about cloud.   Only about 5% were running critical applications in a public cloud.  That goes up to 9% if you throw in non-critical apps.  Private clouds are more in use, with over 40% of users running something in a private cloud, but when we’ve drilled in with end users, private cloud is often just virtualized environment that’s been “rounded up” to a cloud:  There’s no self-service, no abstraction of the server from the end user.

 

We did a separate survey about cloud plans, and with 88 respondents, it told a similar story.  Roughly 70% of respondents had no plans to do anything with cloud in the next year.  Only 16% were planning a cloud initiative in the next 6 months.

 

BTW, if you’re tempted to dismiss these results because SolarWinds is “an SMB player”, let me set the record straight:  Just because we’ve figured out how to sell to customers with only a few hundred employees does not mean that we only sell to that segment.  Our customers—including those in this survey—range in size from hundreds to tens of thousands of customers.  We cover a huge swath of the market.  We just do it without talking to CxOs, who are, perhaps, more susceptible to vendors who “cloudwash” their solutions, given that the CxO can’t easily drill down further than what’s presented in a slide show.

 

Why am I throwing a cold, wet blanket on the cloud party?  Again, we believe cloud is coming, but it’s not here yet, and it probably won’t be here for a while yet, maybe closer to 2020, if IDC is to be believed  (see Public, Private and Hybrid Cloud Adoption – Competing for 2020; IDC). In the meantime, real IT professionals have real problems with their non-cloud environments right now.  And when I look around at the big IT management vendors like VMware, Microsoft, CA, and BMC, they are pushing cloud this and cloud that 7/24/365.  The cloud focus is just as true of startups (although that make sense because startups are all about the future).    Who’s left to look after the problems of today?  That would be us.   SolarWinds continues to focus on delivering powerful, low-cost software that are truly easy to use.

 

We aren’t ignoring cloud.  When it becomes a need for mainstream IT people, we’ll have products that address their pain point.  Count on it.  Until then, if any of our competition wants to pull their heads out of the clouds, we wouldn’t mind a little company in the here and now...

About a month ago, we announced our first release of DameWare since it became part of the SolarWinds family. We are really pleased with the feedback that we have been getting on this release from DameWare customers as well as SolarWinds users who have taken a few moments to check it out.  The DameWare products allow system administrators to easily manage Windows servers, workstations, desktops, and laptops from remote locations. The release brings a couple of great features that the community has been clamoring for, as well as some tweaks and fixes that will make every user’s experience a bit smoother.

 

What’s New in DameWare?

Chat – Now you can chat online with your remote user as you troubleshoot or configure the remote machine. Now, there is no need to open WordPad™ to type back and forth. Just one click on the chat button in the top menu, and a chat window with the remote user opens automatically.

DW MRC Chat Button.jpg DW MRC Chat Window.jpg

Screenshots – Click to quickly capture and save a screenshot from the remote machine. Now there is no need to go through the process of pasting your screenshot into Paint™ and then saving it. One click, one step, and it’s saved! This feature is invaluable in troubleshooting scenarios such as documenting errors or configuration settings, especially since a picture is worth a thousand words!

DW MRC Screenshot Button.jpg

DameWare still has the same super cost-effective pricing model (priced per admin user instead of by managed machine) and all the great functionality you’ve come to expect. In fact, WindowsNetworking.com recently recognized DameWare NT Utilities with two 2011 Readers’ Choice Awards, taking the top spot in two categories -- Administration Tools and Remote Control.

 

Check it out for yourselves!

1 2 3 … 5 Previous Next

Filter Blog

By author: By date:
By tag: