I just noticed I have a set of hikvisions both flapping every Monday about 9:54AM EST. I've never noticed it in a monitoring or recording sense, just happened to notice it when checking switch logs for an unrelated issue. In the screenshots below, ports 36 and 37 are the two camera interfaces. Seems odd. Are the Chinese coming to get me?
Why Is This Hikvision Ethernet Connection Flapping?
The link state is changing from down to up? If so, this looks like an adapter is going bad?
Got it, interesting. Are 36 & 37 on the same switch? If so, can you move one or both to isolate the flapping to a camera/cable issue?
do you have SNMP monitors running on the switch? are you graphing bandwidth/PoE/updown status any way?
Dahua cameras as well as other brands used to be set to reboot daily or weekly. I don't know if Hik has this, but could be an explanation . If the clocks are synced, could explain why they do it at the same time.
...groupies...
Nah, just an unofficial IPVM historian, primarily of the post Rockoff/Boughton period, with an emphasis in Pan-Asian manufacturers...
Have you tested this circuit end to end? We had issues with a new CAT6a cable plant when we used CAT5e patch cables. I learned the hard way how to troubleshoot these types of issues. The circuit would pass a Fluke certifier, but would not pass traffic from a Hikvision camera. We had similar port flapping issues.
Is any part of your data circuit new?
Have you tried using a laptop using the same patch cables and ran iPerf tests back to your server?
Have you tried another camera, different make/model with the same circuit?
I learned the hard way how to troubleshoot these types of issues...
That makes two things in one discussion you learned “the hard way”. ;)
The circuit would pass a Fluke certifier, but would not pass traffic from a Hikvision camera...
That seems incredible, do you remember any specifics, e.g. did it not connect at all, how was it resolved?
I’m surprised you don’t have this story committed to memory with a link to my previous discussion on this subject...
The Fluke tester passed all tests that the electrician (who installed the cable plant) ran. We generally had been using CAT5e patch cables at that point and had no reason to think they wouldn’t work in this application.
The end result was that if we switched to CAT6 patch cables, the issues went away and everything worked fine.
And this story still doesn’t make sense to me either, but it is the truth. Maybe it helps the OP, maybe not.
I’m surprised you don’t have this story committed to memory with a link to my previous discussion on this subject...
Sorry, I must have overslept that day.
Anyway, this statement
The circuit would pass a Fluke certifier, but would not pass traffic from a Hikvision camera.
IMHO, does not seem to comport with that discussion, as the exact circuit that did not pass the Fluke was not the same circuit that did not pass traffic from a Hik.
The 6a plant was only certified from a 6a jack, not thru the 5e patches.
Apologies if I have misunderstood the discussion.
Not that it matters, but the physical connection is from a 6A patch cord direct to switch in same room. These cameras monitor the IDF environment, so there's no cable plant to troubleshoot.
SNMP and graphing statuses and whatnot, I'm sure the network operations people probably have that on a server somewhere.
this goes back a very long time, but I didnt want to upload a 30 foot screenshot of the switch log.
It is every Monday about the same time.
This has to be something the Hikvision system is doing, like a weekly POST or something.
Do any of you have any Hikvisions you can look on the switch and see if yours do a similar weekly thing?
Why not? It’s not like you NEED a CAT6a patch cable for a device that won’t saturate a 100mbps link.
Because when I installed them, I had a box of 100 of 6A (for all the servers also getting installed) and didn't have any 6a/6/5e. And I don't want to fly to Virginia to swap out the patch cord.
Do you really believe scheduled flapping would be stopped by using a 5e patch cord?
There isn’t anyone on site that can swap a patch cable? I’m not sure I would have taken a project so far away if I didn’t have local IT or maintenance on site to assist with such minor tasks.
It is leased space in a colo data center. $150/hr with 1 hour minimums to have someone do it. Plus, they frown upon cameras, they don't know they're there.
Typical integrator answers here. Problem with camera, try swapping the patch cord!
I also said swap cameras. But if you don’t want help, by all means, I will find better use of my time.
I'm so sorry to take up your time. Please, by all means, take care of something more important. Someone out there needs you.
Sounds like the camera is the least of your concern. Here's a good text for you to study.
Typical integrator answers here.
Eh, there's nothing typical about Jon's recommendations/experience. I can't think of any other integrator I've worked with who has heard of iPerf.
He also asked about testing the circuit with a laptop or a different camera which are effective troubleshooting recommendations.
You can't be cynical while at the same time saying you've tried nothing and you're all out of ideas.
Update- I have other hikvisions, same model, in some different network environments. I was able to get those switchports and check the logs, but zero interface flaps on those. Only the two in the OP.
Can you compare the firmware version on the (2) cams in question with the other that don't give you any trouble? Random guess, but maybe the two you have are on a different firmware that have some mysterious Hik related issue.
I don't blame Jon for mentioning the Patch Cord, as many IT related head scratchers can originate from the most basic level. But I cant imagine a Patch Cord with a date and time hard coded into it =)
I wonder if the cable runs to those cameras might be passing by some kind of EMI source? I've heard of issues like this before that were ultimately tracked down to something like a scheduled generator test. Considering this is in a data center which almost certainly has generators, HVAC units and such, it might be worth looking into.
Very plausible, and definitely worth investigating the surrounding area.
Ive got it...Are the cable runs, or the rack housing the switch close to say a break room? And said break room has an ancient microwave that is used to gently warm a very punctual 3rd shift NOC technicians Hot Pockets, during his/her Monday early AM lunch break? Case closed guys.
Don't take microwaves lighlty. I have a friend who was a wireless network installer who was trying to figure out why the wlan kept going up and down. He traced it to a microwave whose shielding was breaking down. As he put it, people were cooking themselves from the inside out.
As he put it, people were cooking themselves from the inside out.
Just in case anyone is not aware, microwaves do not actually cook from the inside out.
I'm sure he was aware of that, but sometimes instead of trying to counter popular thought no matter how incorrect on a technical matter, it's easier just to go with whatever gets across the point "it's not healthy and you need to replace that microwave".
Right on time, yesterday.
Can you confirm Firmware Version of these 2 cams vs the other exact models that dont do this?
Existing versions above...., about to click "Upgrade".
Edit- camera says that is the latest. Hmm. Let me get a screenshot of another camera not doing this.
This one is not flapping. Older version of firmware. They are the same model numbers though, in case that wasn't specified before.
How did this end..?.
It hasn't. It still happens. It isn't causing me any problems so I'm just going to pretend I never saw those switch logs.
Ask questions and get answers to your physical security questions from IPVM team members and fellow subscribers.