You are currently browsing the tag archive for the ‘diagnosis’ tag.
AT&T has just insulted me again
- Issue: DSL starting and stopping (mostly stopping) on a very regular basis (repeatable and capturable with Task Manager and tracert)
- Evidence: started with browser and email “timeouts” (taking too much time to send/receive to completion), videos of any resolution stopping for 15-25 seconds then starting for 5-10 seconds then stopping again, Task Manager showing 5-10 seconds of full bandwidth data followed by 15-25 seconds of absolute 0 bits transmission repeated forever
- What I know: I work with computer systems and do frequent diagnostics of system related problems… I do know quite a bit about DSL, but not nearly enough to do a complete diagnosis but here is what I do know (much of this was after the initial contact while I was collecting data)
- Problem is captured visually with Task Manager and tracert and ping tests
- Problem is repeatable across multiple DSL modems, computers, internal networks
- Problem is not intermittent, it is persistent
- Problem does not happen at neighbor’s AT&T phone line (with my equipment) about 1 mile away (same complex so I would assume same end office)… so it is *not* my equipment problem
- Pattern is regular between on and off, and there is virtually no ‘in between’ (which, in my opinion rules out any line issues which are more chaotic and sporadic)
- Pattern shows distinct points of network failure and distinct points of re-connection (suggesting that some device inside AT&T is having problems keeping a connection with my modems for whatever reason which is the *real* issue that I want to remedy)
- Testing before and during communication with AT&T
- Two DSL modems had the same issues with very similar patterns on my phone line: Motorola 2210 (backup), Netgear combination modem and wireless router (primary)
- Three computers and a smartphone running three different Windows and Android operating systems: Windows 7, Windows Vista, Windows XP, Android 2.3.3
- Both wired (100Mb) and Wireless (WiFi G and N)
- Went to my neighbor’s (about 1 mile away who is arguable on the same end office) and tested my Vista notebook on his network… it worked fine, replaced his DSL modem with my Motorola 2210… my computer worked fine… so the issue was not this computer and this modem
- Collected 20+ window screen shots of: Task Manager networking tab (throughput over time) showing the pattern with different combinations of modems, computers and network media; Tracert to different outside IP addresses that showed one AT&T router that consistently timed out; Ping tests to gateway and other routers showing some packet loss
- Put the issue up on my blog (a previous entry here), and DSLreports forum where one forum participant suggested the solution that AT&T finally came to realize and Twitter and Facebook and others
- AT&T Handling of the matter:
- Standard progression without knowing the expertise of the customer
- Chat w/ Shane on AT&T website: basic test and escalation
- Phone call to discuss network issues escalated to manager Brittany
- Phone call Brittany to advise me of scheduling of tech #1
- Phone call Brittany to me to check to see if tech had arrived
- Tech #1 visit to premises
- Phone call from Brittany on results and schedule tech #2
- Phone call from Brittany on progress
- Tech #2 on premises: Jorge focused on the line quality (an issue but not the pattern problem)
- Phone call to check on progress, found that the case had been closed even after telling Jorge that the pattern still existed and he stated that he would report it to the office. Ted was fixated on my modem being the problem (irrespective of the fact that I had two very different modems and both reproduced virtually the same pattern), and Dave (the manager who had been a network tech) did not want to see what images I had to show.
- Phone call from Dave reporting the scheduling of tech#3
- Tech #3 on premises Ray
- No AT&T person ever qualified me nor asked me if what evidence I had until Ray (who fixed the problem) looked at my data, watched me reproduce the phenomenon, did a ping test, called his office and reconfigured an AT&T router in their network (something that I suggested with *every* AT&T person that I spoke with).
- I estimate that the time involved was: AT&T- 470 minutes (almost 8 hours of actual conversation/time on site… not including any admin or other resources); Customer- probably 16 hours of initial diagnostics, interface with AT&T service, hold, reconnections, data collection, reconfigurations, testing
- at $100/hour for all the AT&T support, that is $783… divided by the $20/mo I spend for AT&T’s Elite DSL service, AT&T will recuperate the cost of this one issue in 40 years (that is 2052), not to mention *my* time and effort
- Suggestions for AT&T:
- Qualify your customer: Neophyte (probably 97% of all customers), Know it all (probably 2%), Helpful (1%) —give us a code where we can bypass all the bozo (‘is modem plugged in? is the computer turned on? we are very sorry for your inconvenience.’)… those that really test and can accurately describe and provide evidence of the phenomenon when asked, those that have technical background and can begin the diagnosis and compartmentalization process.
- Look and listen to what we have to offer… nobody ever asked (until the very last person) what evidence I had, could I repeat the issue
- Use remote access software to view issue from the customer’s perspective (if the network is functioning). Even I use LogMeIn for work on my parents’ computers
- Provide contact information to customer: I never had any way of re-contacting people like Dave with images and data to put in the file (an email would have been very helpful) [calling AT&T always results in 15-30 minutes before actually talking to a person who can do anything… and they are never the same person, so it takes time just to have them read through the file and time to brief them on the important parts that are not in the file].
- Assign one person to champion the customer’s issue through to *resolution*. Most that I spoke with were good, but there was very little continuity and Ted even said ‘we will have to start from scratch’… that’s when I asked to speak to his supervisor.
- Have an internal escalation procedure to get to a real diagnostician earlier in the process if the customer can be helpful [it would save you and me a lot of time]
- Check back after initial resolution to see if the problem is really *fixed*
- Thank helpful customers and even reward helpful customers for their valuable contribution… without any thanks or rewards, customers become irritated… make them part of the “team” … Customers can be a much greater asset if they are not blogging about the non-performing side of AT&T
- Capture information and reward the AT&T employees who actually *fix* the problem with some way of weighting the fixes for their difficulty… and pay/bonus them according to their fixes
Thank you for reading this post to the end…
PS for some reason this blog’s text editor replaces the outline second level bullets with a second level of numbers… quite confusing… my apologies.
My AT&T DSL fluctuates from full throttle to zero in a very regular pattern (using different modems, using different computers with different operating systems)… and having tested one computer and modem at one of my neighbors that is far enough to be on a different path but most probably to the same end office (where it functioned flawlessly)…
Have a look and tell me what you see to be the issue and the resolution… Thanks, J
AT&T technician Ray looked at the data, ran some tests, called the ‘office’, had an AT&T router reconfigured (one in the EastBay) and all is right with the internet world (WoooHooo)…