MikroTik

I've bought 2 units, connected in lab, first one died after ~20h of traffic generator, second one almost one week later but beeing just idle for this time. Both were upgraded to 6.18 just before testing. First one is dead as descibed in other topic, second one is black (no led is lighting), fun is running (rpm differs in time) and there's screechy, hight tone sound from it. Both were connected to crs226 through mikrotik sfp+ copper connector. Now I'm afraid to let it running in my network when I get replacement.

I've bought 2 units, connected in lab, first one died after ~20h of traffic generator, second one almost one week later but beeing just idle for this time. Both were upgraded to 6.18 just before testing. First one is dead as descibed in other topic, second one is black (no led is lighting), fun is running (rpm differs in time) and there's screechy, hight tone sound from it. Both were connected to crs226 through mikrotik sfp+ copper connector. Now I'm afraid to let it running in my network when I get replacement.

Must be bad luck, I have 3 in production since release and still running strong.

We've now had two die on us after only a few days of running.... No lights, no console, no fans, nothing.

Does anyone know the fix for these? I love the box but if they're going to die a week after being turned on..... that makes me nervous.

We have just received 3 of the ccr1009-8g-1s-1s+ for use in our network as we prepare to more our core network to 10 Gig fiber.

All three units we powered on for ~1 week while we preformed standard setup on them (RouterOS v6.19 firmware v3.18), during that time they were not under stress and we didn't notice any problems.

One of the units was put into production, very very simple config, SFP+ bridged with ether1-4. No firewall rules, nothing fancy at all, CPU was always < 4% and traffic fluctuated from ~75 Mbps to ~415 Mbps. This was our first step in evaluating the units.

After less then 48 hrs on a Sunday evening no less, the unit failed. When I arrived on site the unit was off, no fans or anything. After attempting to power cycle the unit to no avail we replaced it with our old Cisco unit to get things back online.

As much as I enjoy doing listening to my phone go nuts with emergency text alerts, and preforming a truck roll on Sunday, we decided it would be best to test the other 2 units we have a bit more before we entered them into production. So today I setup the other two with a simple traffic-generator config between the two units using ether1-4.

The unit acting as the traffic generator crashed with a kernel panic as soon as we started it. Our plan was to upgrade all our equipment to CCR1009 and CCR1036 units, replacing our Cisco and x86 custom routers, but now I am having 2nd thoughts.

Anyone else have any info or advice about these?
I am hoping v6.20 fixes the kernel panic and that I just had really really bad luck with the first unit.

Sounds like there is a flaw in the design. Damn.

Please return all the faulty units to MikroTik, so they could analyse what is the root case of this faults.
...For the sake of other CCR1009 owners

Bookmarking this thread.

Have the exact same thing in mind.Need to replace 2 RB2011's(100d+ uptime) with these for a project but will be patiently waiting until MTK figures out what the heck is wrong with these things.

likely QA issues. there isn't model/brand w/o flaws/bugs or production issues.
but density of flawed units imply need to do something in production/testing to reduced it to sane/common amount/numbers.

I have four 1009's, two 1016's, and nine 1036's in use so far. I've got another 20 1009's waiting for deployment, so we'll see what happens.

It definitely seems to be a QA issue because we have dozens in service with no issues but then we hit a block of three in a row that failed like this then back to another string of no failures...

It definitely seems to be a QA issue because we have dozens in service with no issues but then we hit a block of three in a row that failed like this then back to another string of no failures...

Hi, were those 3 failures in a row from the latest batch (i.e. Sep 2014) ?

Thanks.

Contact your distributor and have them work with Mikrotik on it. Yes there is a problem with a small percentage of CR1009 units and Mikrotik is aware and has a hardware fix ready. But it means a return to implement the fix.

Good practice right now on all CCR1009 is to burn in for 48 hours before deploying until the failed units are discovered and weeded out.

Contact your distributor and have them work with Mikrotik on it. Yes there is a problem with a small percentage of CR1009 units and Mikrotik is aware and has a hardware fix ready. But it means a return to implement the fix.

Good practice right now on all CCR1009 is to burn in for 48 hours before deploying until the failed units are discovered and weeded out.

Thank you. That is very wise advice. What is the recommended method to burn in the device? I have an RB 2011 that I'd like to burn in.

Purchased 8 of these last week.

5 were from one batch (came with 6.13 installed) and 3 were from another (6.18 installed).

Left them all on idling over the weekend, 2 of the earlier batch have failed this morning.

No lights/leds/lcds. Just appear dead. I popped the cover and the PSUs are still supplying 24VDC to the main board, so it appears it's something on there....

On their way back today.

Hopefully the other 6 are stable, or I can figure out how to test them before they go into production.

My one is died after one week of work, and it's been idle before ~20days.
Simptoms are the same, no lights, no fans, nothing, power supply seems to be OK.
Problem started the following, unit rebooted itself, after5-15minits of working, log sad, power outage.
When we arrived the site, nothing worked, no lights, no fans, like the others.
I tried to power up from ETH8, but still nothing.

Tomorow it will be goes back..

In case you see a problem like described here, please contact the seller to arrange a replacement device.

my units are still @mikrotik (almost 2 months is little long IMHO ), but when I'll get replacement, they will get long stress test and I'll let you know about results
BTW, take a look at LCD just hours before they died


https://www.dropbox.com/s/f449rpt5q0k0hx2/ccr.jpg

FYI the units that ship with RouterOS 6.18 do NOT appear to have this problem.

We have them in production and under load and they are going well.

same problem here. 1st unit died after ~10 Day. The replacement only worked 2 days.

Now I've installed an old RB450G

If it wont destroy your warranty, how about trying it with a different PSU of higher wattage? Before it died, did any of you manage to get any temperature readings?

I know this is an old thread, but we just had to replace a faulty CCR1009. It died after ~2 months in production. Replaced it with a RB1000AHx2.

The CCR1009 died with no indication of any problems. No lights, no sound, nothing. We are waiting on our supplier for a replacement.

Snix, please do not raise old unrelated threads. Any device could have a hardware defect. Please just return it to RMA for inspection.