Resolved Frequent short internet outages

I'm using Net Uptime Monitor (Net Uptime Monitor - the simple Internet Connection Monitor) and found multiple short, 5-20 second, outages primarily at night. I switched from U-Verse to Comcast for a number of reasons, and was happy that the outages stopped. Until the first week was over. Outages are less frequent and shorter in duration, but are there. Around the same time I moved to Comcast, I got a new computer. Now I'm still seeing short outages and have all new hardware and ISP. What's going on? Am I concerned over a normal situation? These outages have not interrupted any computer activity, and Netflix doesn't even burp. Since there is no harm, should I be concerned about the foul?

It's curious that there are several Comcast outages occurring at or near 1:12 AM or 12:12 AM. Routine maintenance?

Net Uptime Monitor Failure Log (NetUptimeMonitor.com)
U-Verse outages
=======================================
12/12/2017 5:50:12 PM Log Start
---------------------------------------
Failure Start Length
12/13/2017 4:06:05 PM 0:00:08
12/13/2017 7:23:44 PM 0:00:07
12/13/2017 7:24:09 PM 0:00:05
12/15/2017 1:14:00 AM 0:00:05
12/15/2017 6:13:41 AM 0:00:14
12/15/2017 6:14:09 AM 0:00:09
12/15/2017 6:18:27 AM 0:00:16
12/16/2017 2:40:00 PM 0:00:32
12/16/2017 4:35:31 PM 0:18:09
12/18/2017 1:13:36 AM 0:00:09
12/18/2017 9:12:22 AM 0:02:52
12/18/2017 3:40:18 PM 0:23:40
12/19/2017 5:41:53 AM 0:00:20
12/19/2017 5:42:21 AM 0:00:13
12/19/2017 5:48:22 AM 0:00:11
12/19/2017 6:04:12 AM 0:00:40
12/19/2017 3:30:25 PM 0:05:51
12/19/2017 3:46:32 PM 0:06:23
12/20/2017 9:01:17 AM 0:05:48
12/20/2017 2:20:51 PM 0:00:16
12/22/2017 11:54:12 AM 0:18:14
---------------------------------------
12/22/2017 12:12:27 PM Log End
Monitor Duration 234:22:13
Failure Summary:
Count 21
Total Downtime 1:06:15
% Downtime 0.47
Minimum Length 0:00:05
Maximum Length 0:23:40
Average Length 0:03:09
=======================================

Net Uptime Monitor Failure Log (NetUptimeMonitor.com)
Comcast outages
=======================================
2/2/2018 8:57:44 PM Log Start
---------------------------------------
Failure Start Length
2/3/2018 1:12:22 AM 0:00:06
2/3/2018 10:25:04 AM 0:00:53
2/3/2018 10:39:14 AM 0:00:19
2/5/2018 1:12:12 AM 0:00:07
2/5/2018 2:31:55 PM 0:00:05
2/6/2018 12:23:53 AM 0:00:06
2/6/2018 9:15:30 PM 0:01:31
2/7/2018 2:34:44 PM 0:00:06
2/8/2018 12:12:22 AM 0:00:06
2/10/2018 1:11:59 AM 0:00:06
2/10/2018 3:17:49 PM 0:00:06
2/10/2018 4:36:07 PM 0:00:06

 

Considering several of those reported outages last several minutes (up to 23:40 minutes) and you report no interruption of any computer activity, including Netflix, then I'm with Tony. You are not losing connectivity. Something else is going on. Check your router logs (if your router keeps them) to see if you are losing connectivity there.

Also, if your router and modem are separate devices, replace the Ethernet cable. Those are very critical, but not very robust network devices that are often poorly made and easily damaged. Same with the cable (if not wireless) between your computer and the router.

I would try a different monitoring tool too to see if you get the same results. If you have more than one computer you can run both at the same time with different programs and compare results.

 

Re: 23 minute outage: I can't recall if I was on the computer at that time. Likely not. Netflix buffers their stream, so a 6 second blip would go unnoticed. I'm currently using my own Motorola wireless modem/router (new) and I used the supplied CAT5e cable. It took two weeks with the new computer to realize that internet was defaulted to wireless. But I noticed the same outages wired or not. Since I'm experiencing the same conditions using two entirely different hardware setups and a different ISP, methinks there's something in the monitor program, but I sure don't know what that could be. I'll look for another monitor program. Or I may just forget about it and stop worrying. Thanks to all for responding.

 

Just remember, those supplied cables probably cost $.50 to make and did not go through any testings. It is not uncommon for one or more of the 16 wire crimps per cable (8 per end) to be done poorly. And it takes very little for even a well made cable to damaged.

And the problem is, a damaged cable does not always result in no connection, but rather just poor network performance making it appear the cable is good.

This is why I always make and test my own cables. This way I know for a fact they are good. Plus, I can make them the length I need - if I need a 15.5 foot cable, I don't have to buy a 25ft cable and stuff the extra length into the rat's nest behind my desk.

I am just saying, don't assume any factory made cable was made correctly. And then don't assume any cable that has been used for awhile has not been damaged. It only takes one yank or trip-over to weaken one or more of the 16 "mechanical" connections (crimps). A quality mechanical connection is essential in any connector for there to be a quality electrical connection.

As for finding another program, there are many. Google "free network monitoring software".

And BTW, Netflix does not buffer based on "time". It is on bit-rate. This means the "content" of the stream can greatly affect how much buffer "time" you have. For example, if your content is streaming 720p with 2-channel audio content, you may not see any buffering if you lose your network connection for 6 seconds. But if you are watching 1080p with 5.1 surround sound (or worse, 4K content), you may notice the interruption. Data compression techniques help, but that is dependent on the what is happening in the movie at that point in time. A scene shot in a monochromatic room with just one character moving about, can be highly compressed. While a moving landscape with several moving characters cannot.

 

Thanks, Bill. During my 33 years with MCI, then WorldCom, then back to MCI, and finally Verizon Business (same job, just a different logo on my truck), I've made more than my share of RJ48 connectors, and had to remake a bunch of those. Unfortunately, I was not able to hold on to my crimper when they laid me off in 2011. The tester I do have only shows DC continuity, which is OK, but doesn't guarantee 1 GB transmission. I do have a handful of CAT5e cables, and maybe I'll switch them out periodically just because I can. That's after I try another monitor program. Thanks for the info on Netflix, and your response.

 

That's too bad as I suspect it was a good one.

I learned a valuable and expensive lesson with my crimper, having spent well over $100 for my $45 crimper - not to mention a rapidly receding hairline and elevated blood pressure.

It was not that I was grossly over charged. It was that I was stupid. I started out by buying a $15 crimper. It was a piece of junk. So I bought a $25 crimper. That worked fine for RJ-11 connectors, but when I went to wire my house for Ethernet networking, I discovered it was also a piece of junk. So I broke down and bought the expensive "quality" crimper I should have bought in the first place.

For those wondering how $15 + $25 + $45 = more than $100, once you squeeze down on the crimper, the connector is spent. You cannot reuse it. If you don't get a good crimp the first time, you have to cut it off the cable, throw it away, and use a new connector. Those first two cheap crimpers destroyed a bunch of connectors. And sadly, even with a decent tester (which I didn't buy right away either), you cannot tell which end is bad. So you have a 50/50 chance of cutting off the bad one the first time - assuming the Odds Gods are on your side.

 

You're correct. The crimper the company bought was in a gray plastic box with dies for RJ-11, RJ-45/RJ-48, and the small handset jack. It was a ratcheting crimper. I can't remember the brand, but MCI'S cost was about $120 as I recall. A quick tour through Amazon didn't show me the specific brand. It's one thing I wish I had spirited away when I left, but have only had maybe three occasions to use it in 7 years. Besides, my son-in-law has his own IT company, so I can use what he has if necessary.

 

Found the crimper I used when working. See AMP 2-231652-0 - Crimp Tool, Cutter & Modular Plug Connector USA MADE!!! | eBay

 

I cannot find mine any more but except that mine has red grips instead of yellow, and no brand label above the cutter/stripper blades, mine is exactly like this Fluke.

I've used AMP crimpers before and they were nice. AMP was bought out by Tyco several years ago so not sure if they have maintained the same level of quality.