Where we went wrong was in planting seeds without nurturing them at all. Had we done it right the first time, a day's worth of time would have been dedicated to each page. Had we done that, we'd probably have one of the better resource bases on the Net. Instead, we created the pages, then got caught up in running a business, got traffic from some of those poorly fertilized pages -- then Google kicked us where it hurt.
If you are a victim of the latest Google (Nasdaq: GOOG) algorithm change and experienced an 85 percent traffic drop like the disgruntled fellow penning this article, you may be feeling the heat right now.
It's easy to blame others and point out competitors who are "getting away with worse," but what all of those wounded really need to do is look into the the mirror and figure out how to get better.
Here is a quick guide on how to right the ship if you've keeled over -- or, if you're just starting your website, how to avoid disaster to begin with.
Plant the Seed When It Is Ready to Grow
There is a ton of thinking and planning ahead when it comes to developing a website. In our case, the website was ready before the company officially formed. Oftentimes, this happens in reverse, but the business side of things had to be sorted out.
This led to creating pages ahead of time, with certain "plans" that were to be accomplished soon in regard to content. Most of the pages had some original content, but it was largely generic.
At the same time, we provided consistent updates to our blogs to help the site along for when the business end would be good to go. Google has (or had) a way of rewarding websites that consistently updated their blogs, because it showed fresh content was being added to the mix.
We never copied content, but many of our updates were shallow in terms of coverage. In addition, the blogs were named for the search terms we wanted to rank them for. Not all of them, but many of them were named this way. At the time, it seemed like a good common-sense strategy. Now it looks like a big mistake. We did not monetize off of the traffic at all, and it still backfired.
In sum, create the pages offsite or don't publish them until they are absolutely ready and valuable to readers.
Attacking the Problem
As soon as we realized how Google's algorithm change was affecting our rankings, we went through a checklist of products we wanted to feature and what was still on the back burner in terms of content and resources that could eventually be provided. After that, we went on to delete half of the "resource" pages that had been created.
We then went back to the blog updates and deleted nearly half of those. On Day 1, we deleted probably about 100 pages that were of no real use. Some of those pages were linked to from the home page, yet they offered no real value.
All remaining page guides for the different model phones already had -- or will have -- the proper content and resources in the next few days. After that, we can think about expanding in the never-ending quest to add quality content.
Where we went wrong was in planting seeds without nurturing them at all. Had we done it right the first time, a day's worth of time would have been dedicated to each page. Had we done that, we'd probably have one of the better resource bases on the Net.
Instead, we created the pages, then got caught up in running a business, got traffic from some of those poorly fertilized pages -- then Google kicked us where it hurt.
This is not to say that our competition does not do some lame things that apparently have been overlooked or that Google's system is completely, ahem, fair.
How Long Do We Live in Purgatory?
Right now, we don't even show up for the important local terms, and we don't even have local storefront competition. Many of the websites that clearly don't provide anything of value and fill up the Web with pointless "state pages" rank higher for our local searches.
We still are found on Google, though, and my assumption is that once we have completed many of our resource updates over the next couple of weeks, the new algorithm will reward our efforts.
The obvious question is how long will this take? Our assumption is that it could take one-to-three months. Google is forcing us to go old school, think local and make the website better.
Still, when will it give us the "get out of jail free" card we feel entitled to? After all, if the banks get a bailout surely Google can feel for the plight of a few young and eager repair techs.
Special Treatment
Reports have come out that some website complained and got its rankings back immediately. We don't pretend to think Google master Matt Cutts or the rest of Google give a darn about our 85 percent drop.
Unlike many others who are upset about the crash, though, we understand why we lost favor with Google, and the problem is being corrected.
This is America -- our get-out-of-jail-free card will come. All we have to do is the Google dance, and the search giant will make it rain with rankings, once we've done our penance. With that in mind, maybe we can all start to sleep comfortably
Thursday, March 10, 2011
Using network management tools to catch mistakes of non-engineers
Whenever someone other than a network engineer plugs a cable into an Ethernet port, bad things can happen. A good set of network management tools becomes essential for network engineers who have to sort through someone else's mess.
When the performance of the Extreme Networks switches in his campus network degraded recently, network engineer Brian Saunier looked at his network monitoring tool, Ipswitch’s WhatsUp Gold, to see what was happening.
"When we started looking at it, we could see that the CPU usage was getting high on certain ports and bandwidth was just through the roof," said Saunier, who manages the network at Cobb EMC, a not-for-profit electric utility based in Marietta, Ga.
Using the WhatsUp Gold network monitoring tool, Saunier traced the problem to two specific ports on a core switch. He realized he was seeing the telltale signs of a network loop, with the CPU usage on both ports sky high. A quick visit to the core switch and a little detective work revealed the problem. A help desk technician had plugged a cable into two ports on the same switch, creating a loop. One port on the switch had been pulled to an Ethernet port mounted on a nearby wall. Apparently the help desk tech thought he had found a cable that had been unplugged from the wall, and he simply plugged it back in. A network engineer would have known that the wall port was an extension of one of the ports on the core switch.
Is your portfolio of network management tools complete?
Learn about the network management capabilities of Netflow v9
What network management tools are engineers using on their smartphones?
Asset management is a key element of a network management plan.
Saunier uses several Ipswitch network management tools to keep an eye on what's happening in his network. In addition to the networking monitoring tool, he uses the WhatsConnected network discovery tool and the WhatsConfigured network configuration management tool.
WhatsConnected is a simple network discovery tool that automatically creates a network topology map, offering details on all switches and routers, along with what devices are plugged into each port. By cross-referencing results of that topology map with the records Saunier has of where each port is physically located on Cobb's campus, he can quickly find the physical location of a device. This capability is essential when looking for devices that are plugged in where they shouldn’t be. WhatsConnected has come in handy for Saunier as he prepares his network for a new network access control installation.
"We did a discovery of printers because we were doing a NAC project where we're trying to lock down the network with 802.1x," he said. "We were trying to locate where all the non-authenticated devices were. Printers were our main problem. We did it once or twice during that early phase, and we'll probably have to do it again to make sure nothing else has been plugged in recently."
Ultimately Saunier will apply this knowledge to use Unified Access Control from Juniper Networks to lock down every port that is supporting a printer. .
"We [will] lock down those ports according to the MAC address of the printer so that no other network device could be plugged in there," he said. "If someone came up and said, 'I want to unplug this printer and plug in my laptop,' their laptop is not going to work."
WhatsConnected originally had some trouble with the switches in Saunier's network. Extreme Networks' method for forwarding SNMP information on multi-link aggregation came through garbled, he said.
"Extreme switches do some funky things with their forwarding database, and they didn't release some information through SNMP that Ipswitch needed. It didn't reveal information for [Extreme's] link aggregation groups. When we would run [WhatsConnected], it would look like we had links going everywhere."
Ipswitch provided some quick software customization to deal with the bug, Saunier said. His topology maps have been accurate ever since.
Even network admins can make network configuration mistakes
End users aren't the only ones who can threaten a network by doing something they shouldn't. Even a network administrator can cause trouble by making a network configuration mistake on a switch.
Saunier uses WhatsConnected to audit network configuration changes on his network. If he detects a problem on the network, he can use those audits to track down whether a network configuration change caused the problem.
"It's on a schedule to download configs from all my switches daily," he said. "I can audit it to make sure certain lines are in the configs in a certain way. We're auditing to make sure the RADIUS servers are set on every switch and that the NTP [network time protocol] server is set right on every switch."
WhatsConfigured alerts Saunier if a major configuration change happens or if some other configuration change failed. For instance, if a NTP server configuration change didn't take on a certain switch, he'll hear about it. However, Saunier is careful not to let WhatsConnected overload him with alerts.
“We don't go down into the very granular stuff, like if someone logs in and deletes a port off a certain VLAN,” he said. “I'm not auditing for that because those changes happen so often. When you have a 100-switch network, stuff changes hourly, and I don't really want alerts when every little thing changes."
When the performance of the Extreme Networks switches in his campus network degraded recently, network engineer Brian Saunier looked at his network monitoring tool, Ipswitch’s WhatsUp Gold, to see what was happening.
"When we started looking at it, we could see that the CPU usage was getting high on certain ports and bandwidth was just through the roof," said Saunier, who manages the network at Cobb EMC, a not-for-profit electric utility based in Marietta, Ga.
Using the WhatsUp Gold network monitoring tool, Saunier traced the problem to two specific ports on a core switch. He realized he was seeing the telltale signs of a network loop, with the CPU usage on both ports sky high. A quick visit to the core switch and a little detective work revealed the problem. A help desk technician had plugged a cable into two ports on the same switch, creating a loop. One port on the switch had been pulled to an Ethernet port mounted on a nearby wall. Apparently the help desk tech thought he had found a cable that had been unplugged from the wall, and he simply plugged it back in. A network engineer would have known that the wall port was an extension of one of the ports on the core switch.
Is your portfolio of network management tools complete?
Learn about the network management capabilities of Netflow v9
What network management tools are engineers using on their smartphones?
Asset management is a key element of a network management plan.
Saunier uses several Ipswitch network management tools to keep an eye on what's happening in his network. In addition to the networking monitoring tool, he uses the WhatsConnected network discovery tool and the WhatsConfigured network configuration management tool.
WhatsConnected is a simple network discovery tool that automatically creates a network topology map, offering details on all switches and routers, along with what devices are plugged into each port. By cross-referencing results of that topology map with the records Saunier has of where each port is physically located on Cobb's campus, he can quickly find the physical location of a device. This capability is essential when looking for devices that are plugged in where they shouldn’t be. WhatsConnected has come in handy for Saunier as he prepares his network for a new network access control installation.
"We did a discovery of printers because we were doing a NAC project where we're trying to lock down the network with 802.1x," he said. "We were trying to locate where all the non-authenticated devices were. Printers were our main problem. We did it once or twice during that early phase, and we'll probably have to do it again to make sure nothing else has been plugged in recently."
Ultimately Saunier will apply this knowledge to use Unified Access Control from Juniper Networks to lock down every port that is supporting a printer. .
"We [will] lock down those ports according to the MAC address of the printer so that no other network device could be plugged in there," he said. "If someone came up and said, 'I want to unplug this printer and plug in my laptop,' their laptop is not going to work."
WhatsConnected originally had some trouble with the switches in Saunier's network. Extreme Networks' method for forwarding SNMP information on multi-link aggregation came through garbled, he said.
"Extreme switches do some funky things with their forwarding database, and they didn't release some information through SNMP that Ipswitch needed. It didn't reveal information for [Extreme's] link aggregation groups. When we would run [WhatsConnected], it would look like we had links going everywhere."
Ipswitch provided some quick software customization to deal with the bug, Saunier said. His topology maps have been accurate ever since.
Even network admins can make network configuration mistakes
End users aren't the only ones who can threaten a network by doing something they shouldn't. Even a network administrator can cause trouble by making a network configuration mistake on a switch.
Saunier uses WhatsConnected to audit network configuration changes on his network. If he detects a problem on the network, he can use those audits to track down whether a network configuration change caused the problem.
"It's on a schedule to download configs from all my switches daily," he said. "I can audit it to make sure certain lines are in the configs in a certain way. We're auditing to make sure the RADIUS servers are set on every switch and that the NTP [network time protocol] server is set right on every switch."
WhatsConfigured alerts Saunier if a major configuration change happens or if some other configuration change failed. For instance, if a NTP server configuration change didn't take on a certain switch, he'll hear about it. However, Saunier is careful not to let WhatsConnected overload him with alerts.
“We don't go down into the very granular stuff, like if someone logs in and deletes a port off a certain VLAN,” he said. “I'm not auditing for that because those changes happen so often. When you have a 100-switch network, stuff changes hourly, and I don't really want alerts when every little thing changes."
Subscribe to:
Posts (Atom)