Archive for January, 2008

Is MoneyWell money well spent?

Are you on the following financial plan: Once or twice per month you write out checks/do online bill pay, look over credit card statements (matching them with receipts? Who does that?), and balance your checking account, all while the rest of the month dreading anything to do with your finances?

Yea, me too! But I think I’ve found a way out.

A few days ago I posted about MoneyWell, a new Mac personal finance application that I’m trying out. It uses the concept of Buckets to help you budget your money. It forces you to think of things a bit differently, which may just be the ticket to getting out of debt.

Allow me to expand:

Most any personal finance application can track everything you spend, but the big strength of MoneyWell is in forcing you to create a budget and providing simple tools to help you set goals and reach them. By being reminded how much money you have already spent in a given month on a particular category, perhaps you won’t be quite to cavalier with your spending.

When you first begin using MoneyWell, you must create a Spending Plan. This is essentially a list of categories you are going to spend money in with an estimate of what you think you’ll spend. You must also set precedence for what you need to spent money on first (such as house payment), second (car payment), on down the line. This is usually not that big of a deal as the expenses that are very important tend to take up a large portion of your income, so just stick them at the top with everything else below. Don’t get all crazy on me – It doesn’t have to be exact. In fact, months down the road you can have MoneyWell look at your spending history update it to be better match actual spending patterns.

Next, you have to allocate your income to your spending plan. It automatically fills this in from the highest priority things on down the line, so if you put a category at the bottom of your list for “New Gadgets”, and configure your spending plan as $1000 per month into that category, you’ll only get money allocated to it if all the categories (or buckets) above it get fully funded.

As you add transactions to MoneyWell, you select the bucket that each transaction comes from. This is similar to the categories you select in other applications like Quicken. On one section of the UI (on the main screen – not hidden away!), these buckets are shown along with how much money is remaining in them to be spent or showing how much you’ve overspent. Having this constant reminder of how much you have left to spend in various categories is a very nice feature that helps me stay mindful of my limits.

When a new month rolls around, any overspending or underspending you’ve done in your buckets gets rolled forward to the next month (along with any new allocation). Let’s say you overspent last month dining out to the tune of $50. If you spending plan is to spend $250 a month on Dining Out, when the new month gets here and you’ve allocated your income, your Dining Out bucket will only have $200 in it ($250 – last months $50 overage). If you are consistently overspending in a particular category, you either need to modify your spending plan, or modify your spending habits.
If this sounds difficult to live with, know that MoneyWell is also flexible. If you have money left in another category , you can transfer money from one bucket to another. So, if you had to spend an extra $25 entertaining unexpected relatives, you can move the $25 you saved by skipping a haircut out of your Hair Care bucket over to your Dining Out bucket.

As I mentioned in my previous post on MoneyWell, the way I have chosen to use for maximum effectiveness is to put all my accounts (including credit cards) into it. When I arrive home from a day of shopping (or work – I eat out for lunch a lot), I take the credit card receipts out of my wallet and enter them right into MoneyWell. This might sound like a hassle, but it really isn’t. Previously, I was starting to approach a George Costanza sized wallet with literally 3 months of credit card receipts inside. Now, each evening I enter in the one or two expenses that I had from the day (if any), and I’m done. By looking at this on an almost daily basis, I’m reminded how much money I have left in each category, and I feel like I have my fingers on the pulse of things, financially.A big advantage of tracking your credit card purchases this way is that, once entered the receipts can be dropped into a file folder. When the credit card bill comes in, simply reconcile it as you would any other account. Since you have already entered the purchase data from each of your credit card receipts, reconciliation should be simple.

Checking your credit card statement is something that we’ve all been told we should do. Listen to any consumer advocate radio show (like Clark Howard), and you’ll hear horror stories of companies charging people like $29.99 a month for literally a year or more before they noticed. Once they do notice, the most they can hope to have reversed are charges that took place in the last 60 days.

A side effect of using MoneyWell in this manner is that any unexpected bills or ones that don’t match your amounts will stick out like a sore thumb.

So, yes, it does take me a minute or two per day to manage, but it actually feels freeing, and I don’t think I’m going to dread handling my finances anymore.

“That sounds great and all, but does it really work?”, you are probably asking. I’ve only been using it a week, so I’m sure I’ll be able to say more about it with more certainty in the coming weeks and months. But, I already find myself looking for small ways to save money and limit my spending on a daily basis. I don’t think I’d be doing that if I were still using Quicken. I truly do believe that by keeping tabs on different categories the way MoneyWell does, I will ultimately be in much better control of my money.

January 29, 2008 at 11:45 pm Leave a comment

Cacti Availability Reporting

When I showed off the Cacti administrative interface to the manager of our NOC, I knew I was in trouble.  He astutely spotted the “Availability” column on the Devices screen.  He quickly asked me if he could get Availability reports out of it.  I had used Cacti enough to know that there wasn’t a way to get reports like that, much less on a Monthly basis.

I decided to dig a bit to see what I could turn up.  The availability information along with the two columns that are used to determine availability are right there in the database.  Unfortunately, Cacti doesn’t reset the availability stats, so once a device misses a single poll they are forever doomed to have something less than 100.00% availability.

My management is interested in availability for two things:

1. The department manager is interested in a Monthly Availability report to have something concrete to show the WAN vendor.

2. The NOC manager is interested in a Daily Availability report.  He wants to see a list of sites that have less than 99.95% availability.  He wants to see this in an effort to be able to spot sites that might have minor problems that get missed throughout the day.

Two members of management wanting reports spanning different time spans is problematic.  The availability data is stored in a table in the Cacti database.  My initial thought was that the daily report could be created by simply clearing the appropriate rows in the database once each day, then looking at them the next day just before clearing them.  Doing this, though, wipes out the data for the month, which is important to a higher level of management.

After giving this situation some thought, I came up with what I think is a practical way to handle it.   Each day just after midnight, I’ll read the data from the current day and subtract it from the previous day’s data (that I’ve saved off elsewhere).  It is an exercise in loops and pretty simple logic, really.  That will give me the previous days number of polls and the failed number of polls.  Just perform the calculation to get the availability from my new numbers, then save off the data that I read (the raw numbers) to a file to read in the following day just after midnight.  Finally, on the 1st of the month after doing this, generate the report for the entire month (based on the raw data in the Cacti table), then just clear the fields in the database and delete my history file.

Here’s an example emailed report:

Sites with some bad availability

For anyone else interested, I’ve posted the PHP program I wrote to do this on the Cacti forums, along with a brief description of how to get it up and running on your system:

My HOWTO on the Cacti Forums

January 29, 2008 at 9:58 pm 4 comments

Tracking your finances

I’ve been a user of Quicken for over 15 years now. It has always been a tool that I could count on to help me balance my checkbook, but it hasn’t been a solution that I’ve been particularly happy with, especially since the Mac version that I have (Quicken 2006) doesn’t run natively on Intel Macs. I’ve actually been on a minor quest for something better for a very long time, though perhaps I’ve not been serious about it until just recently.

About a week ago, I bought the MacHeist II bundle. It includes about a dozen applications and two games, including one nice looking tool to handle your finances, Cha-ching. I got the whole bundle for $39 since I had taken part in the MacHeist challenges ahead of time, so even if some of the apps aren’t the best in the world, they were extremely inexpensive.

So, I tried out Cha-ching. It has a very nice looking user interface and a Budget feature that looked really neat. It allows you to set soft spending limits for different categories of spending so you can then go back and see how closely you spent against your goals. It has what I consider a sort of “Mac-ish” feature of being able to attach a photo of something related to an entry. So, if you got a new big-screen TV, you could take a picture of it and store it in your finance app as an icon. It’s an “outside the box” feature, but I imagine it would only be useful if your home was burglarized. Then, perhaps you’d have a picture of your stolen jewelry, right along with information about when and where it was purchased and how much it was.

After starting to use Cha-Ching though, I noticed that one key feature (at least, key to me) is absent. The ability to perform a Reconcile. Yes, that’s right – a personal finance application that doesn’t let your reconcile! I looked around the net a bit, thinking that I had overlooked it, or that perhaps they do it a “Mac-ish” way, but they don’t. You can mark individual transactions as either Upcoming (meaning that date has yet to pass), Pending (not cleared), or Reconciled. But there is no built-in way to check your account entries in Cha-Ching against your monthly account statement. From a bit of reading on the subject, it seems that the developers don’t feel this is a necessary feature. They think that it over-complicates things. I honestly can’t see how they expect their product to be taken seriously as a financial application without this feature. It is not as if the concept is hard to understand, would be difficult to code, or hard to fit into a user interface. I mean, it would be possible for me to go through a process of manually reconciling my account at the end of each month by checking the “Pending” items off as Reconciled one at a time against my statement, but that doesn’t give me the warm comfort of knowing, without a doubt, that my bank’s accounting matches my accounting application, and that I have the amount of money that I think I do. Perhaps the developers are a couple of college students and they have no real background in finance, but this feature is one that should be in any application of this type.

Having said that, I would probably be okay with this potentially hazardous method if I were a bachelor and only had one or two accounts to track. But, just guessing that I have enough money isn’t an option with a family and with the number of accounts that I’m tracking. (You have an ING Direct or EmigrantDirect account for savings, don’t you?? You should!)

In performing my typical surfing, I ran across another new personal finance application that appears to be made by a one-man operation, No Thirst software. It’s called MoneyWell and looks somewhat similar to Cha-ching’s user interface. Big difference #1? It has built-in support for reconciling! The second big difference is that it uses the concepts of buckets to keep track of your spending. You must set your spending plan up before you really use it. This bucket method is similar to the old-fashioned “envelope” method, where people would cash their checks, then split their money up into different envelopes, one for Rent, Groceries, etc. When they ran out of money in a given envelope, that’s it, stop spending on that until next month, or move money from another envelope. Yes, you could just keep moving your money to the appropriate envelope to buy whatever you want, but it does provide a certain definite limit for each category that, if you really are trying to save money, it should serve to curb your spending. Using this method, at the very least, will open your eyes to exactly what you are spending your money on, assuming you do it right.

I watched the MoneyWell tutorial videos and I’m sold enough on the concept that I’m going to give it try. The trial version allows 200 transactions, I believe. I imagine that will be enough to give it a shot for two or three months.

The one glaring feature that is missing from MoneyWell is scheduled transactions, but that is on the developer’s To-Do list. It seems that he has been responsive to end-user requests, so I imagine this will be done pretty quickly.

In my personal situation, we put almost everything we buy on our American Express card, then pay it off each month in full. We do this because it is a CostCo Amex, and we get 1% cash back on everything, plus another 1% back for dining out. In addition to this card, we typically use a VISA when the merchant doesn’t accept Amex. For me to use the MoneyWell system the right way, I must add the Amex card as an account in Moneywell and start tracking each expense that gets charged to the AMEX, assigning each expense to the proper bucket. In fact, I’ve added all of my credit cards to MoneyWell to track this way. It will take more time to manage this, but a few minutes a day should be worth it to keep closer tabs on my spending.

January 25, 2008 at 9:54 am Leave a comment

Sexy Cacti Graphs

I admit it. I’m a newbie when it comes to Cacti. Having said that though, I’ll say that I’m a pretty fast learner, and Cacti is one heck of a tool that can be extended many ways just though a simple WebGUI. I started using Cacti about a month and a half ago, but have only gotten deep into it within the last few weeks.

One thing that has bothered me a bit from the beginning was the somewhat anemic look of Cacti’s graphs. Regardless of how it looks, it does the job of displaying the data, but perhaps the Mac side of me wants it done with a bit more style. So, I set off to finding out a better way.

In the course of updating to the latest version (0.8.7a), I discovered that Cacti now supports opacity, meaning that I can now be able to have area graphs that are partially transparent. Sweet!

However, this wasn’t as easy as flipping a switch. Through literally hours of experimentation, I concluded that setting fairly low settings of opacity was generally not good because the color ended up being too light, making it difficult to see the edge. With mid-level opacity, the colors didn’t seem to turn out to be what you’d expect when they “mixed”… (Yellow and Blue should make a nice Green, not a puke green!) Further, I concluded that just about any color combination with a relatively high opacity setting looks terrible, as it approaches some semi-brown color.

A couple days passed without me giving it much thought. It was as I sat looking at a graph on another website that I realized that this graph looked excellent, even though it had two areas that were very light in color (a very low opacity setting). As I looked at it, I realized the reason that it looked so good was that it had a very dark edge to each of the colors, outlining the highest area. That’s when another idea struck. It was a bit “outside the box”, as it were, but I thought that it would probably work. My idea? Graph the same data twice – Once as an area, and once as a line. Simple – but would Cacti let me do it? And more importantly, would it look any good?

Below is the default look for traffic graphs in Cacti, captured from one of the public demonstration Cacti sites:

Standard Traffic Graph

It gets the job done, but you have to admit that it’s pretty “Plain Jane”.

Now, feast your eyes on the gorgeous graph bits below:

New Cacti Traffic Graph

Now, tell me that isn’t a major improvement in looks?

The secret to the beauty that is above is really quite simple.  In addition to graphing the areas with a level low level of opacity (10% – making them very transparent), I added similarly colored lines set to 100% opacity.  The labels for the areas were removed and placed upon the solid lines.  The spacing between “Inbound” and “Current” near the bottom of the graph is so wide because the areas don’t have labels.

In case I’ve not explained this as well as I could have, here’s a screenshot of the config screen for this Graph Template.  (Click the thumbnail to see it)

Cacti Configuration

Updated 01/20/2015:

I recently had to re-create this and realized I had refined my technique since this post.  I now put the AREA items above the LINE1 items.  The AREAs are added with blank labels, and the LINE1 item is where the label is added.  This helps with the spacing in the graph key.  With LINE1 above AREA, it left a big gap between the key item (Inbound) and the Current: label.  Reversing this eliminates the gap.

January 23, 2008 at 12:55 am 1 comment

Torrents for content delivery over MPLS?

At my place of work, we have an MPLS network to connect our Headquarters site to our retail locations.  Each location is in the process of getting attached to this new network.  One of the problems with having hundreds of remote sites with fairly large pipes (fractional T1 in our case) is that you need absolutely enormous pipes at the central location.

A thought occurred to me today, though, that might make those requirements smaller for “broadcast” files….  You know, files that need to be the same everywhere, such as service packs, videos, music, etc.  What if instead of delivering these large files down to the locations directly from the central office, we could deliver a small torrent file?  That torrent file could direct the remote devices to connect to a tracker at the central site where the file is waiting.  As more clients got the torrent file, they would see each other as places that were available to download portions of the file.  They could connect to each other through the MPLS cloud (without going through the central site) to download portions of the file.  Eventually, the file would be delivered to all of the remotes while the central site would have needed only a fraction of the capacity it would have needed to deliver the large file to each of the sites individually.

With the rise of MPLS, it might be time for the business world to start delivering content this way.  It would save thousands of dollars per month in bandwidth costs at the central site.

January 19, 2008 at 12:23 pm Leave a comment

CactiEZ – Almost an NMS!

Cacti is a tool that I’ve looked at a few times.  Basically, it is a well-organized system for polling a large number of devices and graphing the results.  Recently I found myself looking for a way to track historical data for a few hundred sites.  We looked at CA’s eHealth product, which is pretty amazing, but it is priced at an even more amazing level.  That’s when someone mentioned Cacti to me, so I started digging.  That’s when I found CactiEZ.  It’s a single CD installation, including the OS and everything you need to run Cacti, along with a bunch of plug-ins.

CactiEZ lives up to it’s name!  Several weeks ago I did a quick install of CactiEZ.  It took about 15 minutes.  It’s pre-configured so that all you have to do is set an IP Address, then start adding devices.  I set up graphs for two sites and was pretty pleased with how it worked.  The bad part was all the WebGui work it took to set up a site.  You had to add the device, then add the individual graphs you wanted to monitor for each device.  It was pretty confusing, since I had never worked with it before.

A Hidden Gem.  The Advanced Ping script that I found on the cacti forums looked pretty decent.  I downloaded it and started to install it, only to find that it was also included in CactiEZ, just not loaded.  After loading it, I was pleased with the results.  Essentially, it pings each host several times per minute and graphs the minimum ping time as a green line, and a gray to black gradient to the maximum ping time for that polling interval.  Additionally, if there is any packet loss, it will show up as a vertical band of color behind the plotted lines, the darker the color, the larger percentage of packet loss.

Automation, please!  Ok, sounds good, but I need to add hundreds of devices, but only after they’ve been converted from frame-relay to MPLS.  I couldn’t very well just “keep an eye on it” and add these locations as they were converted.  Then I discovered the “cli” directory.  Inside are some fantastic scripts that let you add devices and define graphs directly from the command line.  It did take me an hour or so running through the various options to figure out (and document) exactly what needed to be done to add a single device and all the graphs we wanted to see, then add it appropriately to the graph tree, but that was the hard part.  Once done, I can simply take a text file containing devices and read it with a simple script that I wrote, then have that script add each device along with each graph, while placing it in an easy-to-use format so end users can get the data they need quickly.

Extra!  Thold is a plug-in that looks very promising.  It allows you to set thresholds for graphs and will alert via email if the threshold is exceeded.  They can be hard thresholds, or based off of a baseline, say one week in the past.  This sounds very useful, but a recent forum post I ran across indicated that the baseline code needed some tweaking.  I’m considering trying to get into the code on this myself and see if I can not only get it working properly for baselining, but see if I can add a feature, such as making it average the last X time periods of data (6 weeks, for example) to create the baseline.  Using an average would probably result in more realistic data, if I can make it work.

Monitor is another plug-in that reminds a co-worker of mine of a product called “What’s Up Gold”.   It basically lets you view a single webpage containing one icon per device that refreshes periodically.  If any device become unreachable, a voice alerts you.  Hover over an icon?  It tells you the availability of that device, along with the last time the device failed.  My only problem was that devices that I added via the CLI tool did not have Monitor enabled by default.  It took a bit of quick coding on my part, but I took care of that with some direct SQL statements.

Cacti looks to be excellent for keeping track of historical data, such as bandwidth utilization, latency, and packet loss.  Give me some more time and I’ll probably be able to make it do even more.

January 12, 2008 at 12:31 am Leave a comment

AT&T DSL/MPD issue fixed

The MPD developers have fixed the issue requiring the dnsnosec option in both Monowall and PfSense that I posted about previously. The fix was only put into MPD5, but Manuel Kasper (lead developer of Monowall) back-ported the fix to MPD 3.18 and now has a pre-release version of Monowall with that fix in it. Once he gets a few reports in verifying that it works properly without the dnsnosec option, he’ll release version 1.233.

So, I don’t imagine it will be very long before the PfSense developers will also have a new version with it included as well.

This is what I love about the open-source community: People working together to overcome issues.  I imagine that if a closed-source product had this issue that they would be happy with a “workaround” (like the dnsnosec option), but the open-source community wants to fix things the “right way”.

January 6, 2008 at 9:54 am Leave a comment

AT&T’s random DSL configuration changes begin!

I’ve been using AT&T (formerly BellSouth) for my DSL service for a few months now.  Sunday night, December 23rd, my DSL connection went down at 8:03 PM.  Of course, I didn’t notice until a bit after 10 PM.  How do I know so precisely when it happened?  I run pfSense on my Soekris net4801 (router), and it has a Link Quality graph that I can call up any time I want to see what sort of latency I’m seeing on my next hop, or if I’m experiencing packet loss to said next hop.

I’m a believer in running PPPoE directly from my router (which means running the DSL modem in bridge mode).  I feel like I have a better view into what’s happening this way.  Perhaps I’m just a control freak.

My first troubleshooting measure was to power fail the modem and reboot my router.  What could it hurt, as I was already down?   Didn’t help.

So, I moved the cable from my Mac to my DSL modem so I could get to the config pages and reconfigured my DSL modem to perform PPPoE.  I did that, and my Internet connection was up and running.

Thinking that perhaps the outage that I was experiencing had just ended, I reversed the process, setting it back to bridge mode, hooking my PfSense box back up, and waiting a bit for PPPoE to establish itself.  I waited, but it didn’t come back up.

I reversed the process again, going back to PPPoE mode on the DSL modem, and I was up and running again, immediately.

Is this some sort of plot by AT&T to ensure that only AT&T hardware can connect via PPPoE, I wondered…  So, I tested that theory…  After changing my DSL modem BACK to bridge mode, I left my Mac attached to it, and set it up for PPPoE, ensuring that I copied the username/password directly out of PfSense (which had been working fine until 08:03 PM).  Up and running!  So, my Mac could do PPPoE direct.  I ran a few speed tests and all seemed fine.

I moved the PPPoE back over to PfSense, but there was still no joy.

Stumped at this odd situation, I figured that instead of wasting any more time that night (it was about midnight by this point), I’d post my experience on the PfSense forum to see if anyone else had ran across this issue.  I ended up running with PPPoE on the DSL modem with PfSense set for DHCP mode.

The next morning, no reply, so I simply added a few lines regarding the fact that my Link Quality graph wasn’t much use, since it would now be monitoring a next hop of my DSL modem.  (I don’t expect my patch cords to go bad very frequently.)  By that night, the lead developer replied, pointing me to a post that allowed me to set the address that gets monitored by the Link Quality graphs, enabling me to continue running in DHCP mode, but with a useful graph.

Many days go by and I happen across the forum again, only to see that numerous others are now experiencing the very same PPPoE issue that I had experienced.  Other forum members start digging into the problem, but things seem slow going, so I investigate their findings and start looking at source code…

It turned out that AT&T had modified their configuration to only hand out a single DNS server.  MPD is the PPPoE client in PfSense and it contains a bug which causes MPD to believe that the IP Address field was being rejected by the PPPoE server, when it was rejecting MPD’s request for a secondary DNS server.

After a bit of back-and-forth on what we should do, the lead developer pipes up again about a hidden configuration option which configures MPD to not request a secondary DNS server.

To see the specifics on the hidden config option, go here.

If MPD doesn’t ask for a secondary DNS server, it can’t get rejected, so the “buggy” code never gets hit.  That turned out to be a good short-term fix for this issue, as I’m up and running now on PPPoE again.

Of course, no one at AT&T could even admit that some configuration change actually had happened.

Note that it looks as if AT&T is still rolling this change out, as people keep coming forward every few days on the PfSense forum stating that they have also ran into this issue.

I’m guessing that this has something to do with their horrific DNS server situation.  The DNS servers that AT&T inherited from BellSouth have been the culprits behind numerous issues that my employer has encountered.  If your machine is configured with multiple DNS servers, it will communicate to all of them to resolve every DNS request.  It takes the first one to reply, and runs with that address.  If each customer is only handed one DNS server, requests from each customer won’t be hitting a pair of servers, but rather a single server, so each server should have less traffic if they split it up properly.  This is a short-sighted fix, though, since the end result will be that the one DNS server that a customer is handed is suddenly much more important to be running all the time.  In this case, a failure of one DNS server won’t partially effect a really major number of customers, but rather would have the effect of completely taking down the internet connection (essentially) for all the customers unlucky enough to talk to that DNS server.

Me?  I use OpenDNS.com, so AT&T can do whatever they want with their crummy DNS servers.

January 5, 2008 at 12:05 am 1 comment

Troubleshooting Internet Latency & Packet Loss Issues

I got an email today asking for advice troubleshooting cable Internet issues. They seem to think they were suffering from intermittent latency issues. As I was writing my reply, I thought others might benefit from this as well, regardless of if they were using Cable or DSL.

I have had issues with both Cable and DSL in the past. About the best free troubleshooting tool that I can recommend for people who aren’t overly technical is UOtrace. You can get it here:

http://support.uo.com/latency.html

It’s a Windows tool (sorry Mac users, I’ve not been able to find an equivalent) written specifically to troubleshoot latency issues with a game called Ultima Online, but you can put any server or website name you want in the field at the top (ignoring the ones in the drop down list). It may be best just to test to their servers, though since we know they are allowing ICMP traffic all the way through their firewalls. Here are some quick instructions (adapted from the website link above):

  1. Double-click on UOTRACE.EXE to run the program.
  2. Click on OPTIONS, then ADVANCED.
  3. Choose the server you wish to trace to from the drop-down list. (This is where you can also just enter the name of any site you want to use as a test endpoint.)
  4. Click the TRACE ROUTE button. A succession of host names will begin to appear in the main window. These are the computers your data passes through on its way to the desired server. Once the function is complete, a message that says “Traceroute successful” will appear in the lower left corner of the window. If you get a message saying “Maximum number of hops exceeded” click on OPTIONS then SETTINGS and change the max hops to a higher number and re-trace.
  5. Click the POLL button. This tells the program to send multiple packets of data through the route to the final host. As the data is being sent, look at the PKTS R/S column. Let it send about 100 packets then select STOP POLL.
  6. If you want to copy the results to send to someone else or post on a forum, click Copy in the Edit menu, then paste it where you want it to go.

Once you have stopped the polling, take a look at the results. The list displayed there shows every hop that your machine goes through to get to the destination server. Now it’s time to interpret your results…

  • When you are experiencing latency issues, you’ll probably see that your first few hops are fine, with very low round trip times, but then at some point you’ll hit a router that is responding much slower than the previous ones. You’ll probably notice that all the hops after this point have high latency. That hop, or the one just before it, is the problem. You can get this data to someone at your ISP, and if they can filter it up to their network department, they should be able to take a look into the problem.
  • If you see packet loss to every hop past a particular hop, the first hop with loss is probably the problem. (All packets to hops past the first one with loss have to pass through the one that is dropping packets.)
  • If you see packet loss pretty much from start to finish, it’s probably an issue with your cable/DSL modem, or possibly your connection itself.

Notice that I’ve started talking about packet loss. That’s because most people aren’t able to tell the difference between latency issues and packet loss issues. You see, TCP/IP handles dropped packets gracefully, and will retransmit if it doesn’t get a reply within a certain time span. Perhaps on the second transmission, your packet will get the proper reply, so it may just look like things are running slower than usual. In this case, latency isn’t the problem at all, but rather packet loss.

Note that weather can be a factor as well. If the connection works fine most of the time, but you start seeing latency issues on rainy days, or when it gets extremely cold or hot, then chances are that it isn’t really a latency issue, but a packet loss issue. I’ve seen issues where problems have cropped up after having a lot of rain, then it would clear up later after things dried up. The problem here is that you can’t get a repair man onsite when the problem is happening. In some cases, you might end up with a tech who would pretty much look at it and say “It’s working now”, and not want to really try to do anything to fix it. In other cases, you might get someone who will replace a splitter, just to be able to say they replaced something. I was lucky enough to get a tech with the right gear to diagnose the problem. I had a marginal signal from the ISP. When I had a lot of rain, the signal degraded to the point that I had severe packet loss. He was able to make changes that boosted the signal strength and that fixed my issue.

Another problem that comes to mind from my experience years ago was between my ISP and another upstream provider. This was problematic because the two sides wanted to point fingers at each other. I think they got it fixed though, eventually.

Of course, I’ve since cancelled my Cable internet service, switched to DSL, then later switched back to Cable, and now I’m back on DSL, since they started offering Naked DSL (DSL without requiring phone service) in my area.

Intermittent issues can be very frustrating to troubleshoot, especially if your ISP isn’t very responsive. Hint: That’s pretty much all of them! So, just try to remain calm.

Above all, if it gets to the point that you have a service person sent out from the ISP, be very kind to them and explain the issue in as much detail as possible. Realize that they probably deal with a few intermittent problems every week, and it can be very frustrating to them also, when they can’t see the problem.

January 4, 2008 at 11:19 pm 3 comments


Calendar

January 2008
S M T W T F S
 12345
6789101112
13141516171819
20212223242526
2728293031  

Posts by Month

Posts by Category