http://www.mikrotik.com/download

What's new in 5.20 (2012-Aug-15 13:04):

*) manual upgrade to NEW beta poe controller firmware v2.0 for RB750UP and OmniTIK UPA;
more info at http://wiki.mikrotik.com/wiki/Manual:PoE-Out
*) fix RB951-2n wireless issues;
*) ups - fixed resource leak;

uldis
PPP CPU Load ?

Still have an issue with reboot of RB751 - Ticket#2012080166000114] RB751U-2HnD - Reboot

Now that's funny. NV2 nothing. Migrate to the Airmax UBNT. Clients will not wait

Now that's funny. NV2 nothing. Migrate to the Airmax UBNT. Clients will not wait

I hope for NV2 changes, but nothing.

What's new in 5.20 (2012-Aug-15 13:04):
*) manual upgrade to NEW beta poe controller firmware v2.0 for RB750UP and OmniTIK UPA;
more info at http://wiki.mikrotik.com/wiki/Manual:PoE-Out

Bravo! seems to work pretty good here.

very good for here!

• Can't seem to get the UPS to work from the USB port on a RB751G using 5.20...
  Connected ok under 5.19 but would not stay online longer than 5 minutes even if I set it to 60 minutes. UPS was online and ok.
  The UPS is a APC RS 1500.

See the image below....

Very good.
problem solved at RB411AH + R52HN.
thank you very much.

regards
lazuardy

Very disappointed that no SSTP fixes made it into this release.

Very disappointed that no SSTP fixes made it into this release.

Have you actually tried it?

Very good.
problem solved at RB411AH + R52HN.
thank you very much.

regards
lazuardy

Hello Lazuardy! Which problem do you mean?

Now that's funny. NV2 nothing. Migrate to the Airmax UBNT. Clients will not wait

Could anyone please remind me what we're specifically talking about?

Very good.
problem solved at RB411AH + R52HN.
thank you very much.

regards
lazuardy

Hello Lazuardy! Which problem do you mean?

ya, normis
I previously had trouble RB411ah / R52HN hangs after 6-7 hours at ROS v5.19, Downgrade v5.18 running but ccq ccq value is very small 20 % - 40 %, data rate and throuhput not good.
After upgrading to 5:20 RC1 and v5.20 released hotspot connection is stable.
The value of data rate, throughput is very good, it was ccq average 80% -90%.
Nice job

RB411ah / R52HN AP bridge only
I using RB450G router.

it´s joke? no!

Show column Distance

real max 1km.
mikrotik distance max 3km, average 1km.

Upgrading to the new POE firmware (previous 1.10, and new 2.0) on RB750UP got problemes on ether5 when connecting a Rocket M5 to it.

With PoE firmware 1.10 it works well, but upgrading it to 2.0, it continously reboots de Rocket M5. We try it with two RB750UP units, and does the same.

If set "PoE Out" to auto on ether5 (nothing connected in it) and ether4 (Rocket M5 connected in it), the Rocket M5 reboots faster. And when set "PoE Out" in ether5, Rocket M4 on ether4 works fine.

What if you set it to "forced on"?

I seem to be getting better CCQ for NV2 in PtMP configuration. Can Mikrotik confirm that there were changes made to NV2? I'll not deploy until someone confirms it. Thank you.

Minimal improvements have been done to the AR9344 Atheros driver, but mainly for RB951 and RB2011. It could have some effect on other AR9344 devices.

What if you set it to "forced on"?

"Auto on" or "Forced On" the same result.

Trying to power the Rocket M4 in ether4, with "auto on" or "forced on" at ether5, makes power at ether4 reset faster. And if I unplug de RocketM5 from ether4, red light from ether4 blinks until I put ether5 to "off". And the "PoE Out Current" shows 1920mA in ether4.

After that, the RB750UP rebooted and, unplugging the RocketM% from ether4 and anything plugged at ether5, with "forced on" in 4 and 5 it shows 1900mA and red light blinking.

RB411AR with an R52n. The R52n seems fine, but I can't choose turbo on the built in 2.4GHz card.

??

When did we lose 40MHz for A/G devices?

Here we go again, changes to critical components and no mention of it in the changelog...

lost my statics :S

any with the same problem?

what about the RB1100AH, RB1200 port labelling issue?

Here we go again, changes to critical components and no mention of it in the changelog...

Which changes do you mean? The only changes are the ones in there.

I seem to be experiencing similar issues as "sindudas" with the POE after upgrading.
I went from v5.18 with the 1.x firmware to v5.20.
Tried with Ubiquiti Picostation M2 (Unifi firmware), as well as Ubiquiti AirCam.
Both devices worked without issue on v5.18 w/1.x FW.
Seeing constant resets on v5.20 w/2.0 FW.
Tried 'auto on' and 'forced on' no change.
Adjusted priority from 10 to 0. no change.
Adjusted prority from 0 to 99. no change.

/interface ethernet poe settings> print
and
/interface ethernet poe settings> export
as follows.....

[puterboy@MikroTik] /interface ethernet poe settings> print
version: 2.0
ether1-poe-in-long-cable: no
[puterboy@MikroTik] /interface ethernet poe settings> export
# aug/16/2012 21:44:10 by RouterOS 5.20
# software id = G602-XFV4
#
/interface ethernet poe settings
set ether1-poe-in-long-cable=no


Anything else I can try to help sort this out?

Thanks!
~puter

Hello Folks!

Tonight we upgraded all our production RB333, RB450G, RB750, RB411, SXT-5D and RB600 devices from 5.19.

No customer impact and no problems at all so far and our network is heavely loaded most of the times, many ppl working from home with citrix and other critical stuff.

Again, good work mikrotik!

Hello Folks!

Tonight we upgraded all our production RB333, RB450G, RB750, RB411, SXT-5D and RB600 devices from 5.19.

No customer impact and no problems at all so far and our network is heavely loaded most of the times, many ppl working from home with citrix and other critical stuff.

Again, good work mikrotik!

Good to read upgrading of your entire network to v5.20 was a success but I could not take a similar approach to upgrading live networks, I have upgraded just one less critical PTP and testing so far has been positive then I will move slowly but steadily through the network, remember customers who have a choice of providers have zero tolerance to interruptions to their service.

RB951-2n still have problem with wireless
But CPU now working on 400Mhz.

RB951-2n still have problem with wireless
But CPU now working on 400Mhz.

What problems are you having? maybe they are related to the environment or configuration?

Make sure you use AES and not TKIP, change "distance" to "indoors", make sure you have b/g/n not only-b, try other frequencies.

RB951-2n still have problem with wireless

I'm not seeing that. Installed my first one yesterday. Client connects with an ipad and is getting great speed up and down.

Very disappointed that no SSTP fixes made it into this release.

Have you actually tried it?

No, but there was nothing in the changelog.

Excellent job on wireless! Every single AP(RB411, RB600, RB435, RB433, ...) - PtMP and also PtP are now stable.
We had many "reassociatings" on 2.4GHz 802.11b with old 2.4G APs, but now is everything OK.


Perfect!!!!!

When i try to add UPS - The following errors: "Couldn`t add New UPS - opening serial port failed: 2 9 (6)"
5.19 - all OK

just took a nice shiny new 1100AH out of the box and loaded 5.20, all configured nicely deploying tonight. don't anticipate any problems have great faith in Mikrotik products and software. will obviously report here if any occur. Its replacing a 1200 running 5.16 that is slightly under-powered for its job.

just one slight niggle I would have liked to seen the internal Fan pinouts on the board so I have the option to add a Fan if it got a little hot under the bonnet. It just has holes where the pinouts would go.

Can anybody at Mikrotik tell me. are those holes actually connected into the power circuit allowing me to solder in some pins if I wanted to?

5.20 uptime 2d routerboard RB433UAH (power 24V 2A) problem not responding routerboard (stop traffic, no winbox,no telnet,no wirelles, ether,no SMB, no graf statistic KO! linux) (-//-random stop v.5.18;5.19) >> power off-on = OK functional again (watchdog disable)

RouterOS 4.xx - 5.?? uptime from version to new version

When i try to add UPS - The following errors: "Couldn`t add New UPS - opening serial port failed: 2 9 (6)"
5.19 - all OK

Same here, on 5.19 all good.

5.20 uptime 2d routerboard RB433UAH (power 24V 2A) problem not responding routerboard (stop traffic, no winbox,no telnet,no wirelles, ether,no SMB, no graf statistic KO! linux) (-//-random stop v.5.18;5.19) >> power off-on = OK functional again (watchdog disable)

RouterOS 4.xx - 5.?? uptime from version to new version

I had same issue. turned out to be the dreaded green cap syndrome

RB1100AH humming away nicely on 5.20.
Just one minor issue so far:

I have DUDE monitoring CPU temp using CPU Temp [oid("1.3.6.1.4.1.14988.1.1.3.11.0")/10 (the /10 was a fix for the 1200)

It now shows The second ICE age has occurred over night reporting a temp of -273.0

Mikrotik can you tell me how to correct the readout thx

Now that's funny. NV2 nothing. Migrate to the Airmax UBNT. Clients will not wait

Still, I upgraded to 5.20. And I must say that it is Nstreme NV2 and more stable!
Although nothing in the changelog.
In some places, the stability improved. In some places, has not improved. Anyway, it's a step forward! Thank you.

Sorry for my English.

Upgraded omitik upa and a lot of Device lost radio interface!

Upgraded omitik upa and a lot of

Code: Select all

system,error,critical System rebooted because of kernel failure

Device lost radio interface!

Have you latest routerboot?

Upgraded omitik upa and a lot of

Code: Select all

system,error,critical System rebooted because of kernel failure

Device lost radio interface!

Yesterday I was one click away from upgrading one of our UPA's when I saw some client traffic moving and postponed upgrade Now it kinda scares me to do it

Issue persist:
RB1100AH ports name via Terminal, SNMP and Physical

Issue persist:
RB1100AH ports name via Terminal, SNMP and Physical

I have RB1100AH & RB1200 I am not seeing this. But then I did configure these routers from scratch.

I did not upgrade a pre configured router.

Now that's funny. NV2 nothing. Migrate to the Airmax UBNT. Clients will not wait

Could anyone please remind me what we're specifically talking about?

Talking about this
Fortunately, I have set the client only 1 frequency AP. (scan list) To recconect took the least time.

RB1100AH humming away nicely on 5.20.
Just one minor issue so far:

I have DUDE monitoring CPU temp using CPU Temp [oid("1.3.6.1.4.1.14988.1.1.3.11.0")/10 (the /10 was a fix for the 1200)

It now shows The second ICE age has occurred over night reporting a temp of -273.0

Mikrotik can you tell me how to correct the readout thx

This seems unique to the 1100AH. I have now upgraded a RB1200 and the CPU temp is reported correcty

This night, I upgraded about 10 devices. One of them I was not able to connect after upgrade....When I came to place, wireless interfaces havent been in routeros and also no wireless packages....it was rb433 and before it was there ros 5.18.....So be careful, what If I had this device in tower with bad access.....

what If I had this device in tower with bad access.....

http://www.gsmcommander.com

I'm also rolling out 5.20 over hundreds of Routerboards. So far I've only had a problem with one RB750 not restarting. The GSM Commander saved the day as the RB750 is very far from where I am sitting.

That package wasnt there, the only way how to solve it is to copy it there again or using netinstalleter. So this gsm commander can be good, but not for this case....

3000 dollars for gsm commander? Are you joking me? For that money I put much stable devices there......

Prices are in South African Rand $1 = ZAR8.33

current version firmware?

RB1100AH humming away nicely on 5.20.
Just one minor issue so far:

I have DUDE monitoring CPU temp using CPU Temp [oid("1.3.6.1.4.1.14988.1.1.3.11.0")/10 (the /10 was a fix for the 1200)

It now shows The second ICE age has occurred over night reporting a temp of -273.0

Mikrotik can you tell me how to correct the readout thx

a little further investigation reveals my RB1100AH does not have a CPU temp sensor. Not a major problem for me but the sales brochure says it does. does my board have a fault or is it just a typo in the sales brochure?

RB1100AH humming away nicely on 5.20.
Just one minor issue so far:

I have DUDE monitoring CPU temp using CPU Temp [oid("1.3.6.1.4.1.14988.1.1.3.11.0")/10 (the /10 was a fix for the 1200)

It now shows The second ICE age has occurred over night reporting a temp of -273.0

Mikrotik can you tell me how to correct the readout thx

a little further investigation reveals my RB1100AH does not have a CPU temp sensor. Not a major problem for me but the sales brochure says it does. does my board have a fault or is it just a typo in the sales brochure?

Mine shows temperature and power usage -ROS 5.14

Statistics quickset stopped. the same situation is with ros6b3

Make sure to upgrade to final v5.20 version first. It's possible that RC1 had this issue, but Final doesn't. http://www.mikrotik.com/download

demo2.mt.lv of ROS on 5.20 also is blocked

RB1100AH humming away nicely on 5.20.
Just one minor issue so far:

I have DUDE monitoring CPU temp using CPU Temp [oid("1.3.6.1.4.1.14988.1.1.3.11.0")/10 (the /10 was a fix for the 1200)

It now shows The second ICE age has occurred over night reporting a temp of -273.0

Mikrotik can you tell me how to correct the readout thx

Please write to support@mikrotik.com your board Serial Number so we could check and try the board from the same batch.

RB1100AH humming away nicely on 5.20.
Just one minor issue so far:

I have DUDE monitoring CPU temp using CPU Temp [oid("1.3.6.1.4.1.14988.1.1.3.11.0")/10 (the /10 was a fix for the 1200)

It now shows The second ICE age has occurred over night reporting a temp of -273.0

Mikrotik can you tell me how to correct the readout thx

Please write to support@mikrotik.com your board Serial Number so we could check and try the board from the same batch.

Hi
I have sent the email this morning. Curiously my board temp sensor also stopped this morning. so just in case its a software issue I also included a SUPout file.

Thanks for your attention

5.20 uptime 2d routerboard RB433UAH (power 24V 2A) problem not responding routerboard (stop traffic, no winbox,no telnet,no wirelles, ether,no SMB, no graf statistic KO! linux) (-//-random stop v.5.18;5.19) >> power off-on = OK functional again (watchdog disable)

RouterOS 4.xx - 5.?? uptime from version to new version

-//- routerboard KO! >>> power off-on = OK functional again

5.20 uptime 2d routerboard RB433UAH (power 24V 2A) problem not responding routerboard (stop traffic, no winbox,no telnet,no wirelles, ether,no SMB, no graf statistic KO! linux) (-//-random stop v.5.18;5.19) >> power off-on = OK functional again (watchdog disable)

RouterOS 4.xx - 5.?? uptime from version to new version

-//- routerboard KO! >>> power off-on = OK functional again

we have the same problems

I have a RB751G-2HnD with 5.17 software version.

When i upgraded to 5.20 and firmware to 2.41, I've noticed that every 30 to 60 seconds there are 15-20 sector writes.

I've accumulated 200 000 sector writes in a couple of days.

Then I downgraded to 5.17 and the sector writes didn't improve. When i downgraded the firmware back to 2.38, everything returned to normal. I now have 154 sector writes since reboot and that was an hour ago (this is a normal value after reboot).

Is anybody else having the same issue?

I was watching writes for a while and here are results
ROS 5.20 was installed at midnight, there are some file uploads and config changes before 21:30 So i have more than 400 sector writes/hour without any config changes or logging. Graphs are stored every 24 hours.
Total writes 340 644 / 13 days (all with firmware 2.41 and ROS 5.19 until today).

RB751G with OS v5.20 and firmware 2.41

I have a RB751G-2HnD with 5.17 software version.

When i upgraded to 5.20 and firmware to 2.41, I've noticed that every 30 to 60 seconds there are 15-20 sector writes.

I've accumulated 200 000 sector writes in a couple of days.

Then I downgraded to 5.17 and the sector writes didn't improve. When i downgraded the firmware back to 2.38, everything returned to normal. I now have 154 sector writes since reboot and that was an hour ago (this is a normal value after reboot).

Is anybody else having the same issue?

http://wiki.mikrotik.com/wiki/Manual:Ro ... bad_blocks

Since RouterOS v5.18, NAND is refreshed every few minutes which increases sector writes by ~32 per refresh. At first this may seem very much but it is implemented that refreshing writes 1 write cycle to NAND per week. As the manufacturer of the NAND chip guarantees 100'000 write cycles to the NAND sectors, NAND life is about 100'000 weeks.

Is this really a issue or just the way MT have refreshed sector writes and should we be concerned by this?

abucan, this is the v5.20 topic, but you run v5.17.

in the next version, this issue with NAND writes was fixed, so please upgrade to v5.20

v5.18
*) fix bad block count not to increase on Samsung K9F1208U0C nand;

abucan, this is the v5.20 topic, but you run v5.17.

in the next version, this issue with NAND writes was fixed, so please upgrade to v5.20

v5.18
*) fix bad block count not to increase on Samsung K9F1208U0C nand;

I just wanted to point out from witch version I did upgrade to 5.20. I have no problem with 5.17.

As I can read from Wiki, from version 5.18 sector writes are a feature and not a bug "Since RouterOS v5.18, NAND is refreshed every few minutes which increases sector writes by ~32 per refresh."

I didn't know that.

Thanks.

Sorry if I misunderstood you. We have updated the text to be more readable: http://wiki.mikrotik.com/wiki/Manual:Ro ... bad_blocks

bug in sstp??

Ticket#2012082266000094

RB433UAH CPU 100%



Something is prepared for metarouter ?
metarouter.net
metarouter.com

Mikrotik Support

Please pay urgent attention to the following support incidents:

2012082266000638
2012082066000356

I think I may have discovered a bug in 5.20 802.11 2.4GHz wireless.

We are having mixed results with POE ports on rb750UPs since we upgraded to 5.20.

Basically we upgraded a number of rb750ups setup with a bridge between ports ether1 and ether2 as a master port. Effectively we are using them a POE power supply/switch with the ability to reboot connected APs.The firmware, the routerboard and the poe firmware were all upgraded. All seem fine given that they have only been upgraded a few days ago.

The other popular configuration we use is to separate ether5 from the ether2 master group and power the radio feed (either a Mikrotik or a Ubiquity unit) off that port. We then configure routing between ether2 and ether5, vpns etc. Sometimes we power a local wifi access point from one of the ports ether2, ether3 or ether4. Of five rb750UPs in this configuration when updated four don't maintain poe on port ether5 to the radio. All had been running reliably before. The quick fix was to insert a POE injector between ether5 and the radio.

We took two unused rb750UPs, updated them by updating the main firmware, then the routerboard and finally the poe firmware. We set them up ready to replace a pair of the problem units and tested the poe on port ether5 with a Nanostation. All looked good until we tried replacing the problem routers.

Does anyone know how the poe sensing really works? I am wondering whether we may have a cabling issue that doesn't cause problems with a poe injector or with the earlier firmwares, but does now. All the affected customers are running vpns between offices and won't tolerate an outage on a system that is running to satisfy our curiosity.

We are having mixed results with POE ports on rb750UPs since we upgraded to 5.20.

Basically we upgraded a number of rb750ups setup with a bridge between ports ether1 and ether2 as a master port. Effectively we are using them a POE power supply/switch with the ability to reboot connected APs.The firmware, the routerboard and the poe firmware were all upgraded. All seem fine given that they have only been upgraded a few days ago.

The other popular configuration we use is to separate ether5 from the ether2 master group and power the radio feed (either a Mikrotik or a Ubiquity unit) off that port. We then configure routing between ether2 and ether5, vpns etc. Sometimes we power a local wifi access point from one of the ports ether2, ether3 or ether4. Of five rb750UPs in this configuration when updated four don't maintain poe on port ether5 to the radio. All had been running reliably before. The quick fix was to insert a POE injector between ether5 and the radio.

We took two unused rb750UPs, updated them by updating the main firmware, then the routerboard and finally the poe firmware. We set them up ready to replace a pair of the problem units and tested the poe on port ether5 with a Nanostation. All looked good until we tried replacing the problem routers.

Does anyone know how the poe sensing really works? I am wondering whether we may have a cabling issue that doesn't cause problems with a poe injector or with the earlier firmwares, but does now. All the affected customers are running vpns between offices and won't tolerate an outage on a system that is running to satisfy our curiosity.

Please give us some more info:

1) whether port 5 does not power-on at all?
2) whether forced-on mode helps?
3) output of
/interface ethernet poe print
/interface ethernet poe monitor [find] once

bug in sstp??

Ticket#2012082266000094

Ongoing.

Happens a lot on my RB1000 w/ ~200 clients. Other times, the SSTP service just bombs out and nobody can connect until its restarted. Been waiting on a fix for a looong time.

Same here. v5.19 UPS package works fine, but not in v5.20. Put everything back to v5.19 and UPS is good and working. Need this fixed!

Thanks!

Brad

When i try to add UPS - The following errors: "Couldn`t add New UPS - opening serial port failed: 2 9 (6)"
5.19 - all OK

There is a bug in 5.20 bridge !!!
I checked several boards running 5.20 and all of them show strange results in bridge host table.
i.e.
this is from a 750G 433AH RB333 again running on 5.20 Mac addesses from interfaces not belonging to a particular bridge, appear in its host table.
Otherwise things seems to be working ... no loss of connectivity. But.... ?
Anyone else ?

i just upgraded m RB-750GL and Rb433 to V5.2 and no issues

Pavlan how many ports do you have on your bridges?

Any update on the UPS Package being broken in v5.20? You can select usb1, but it hangs and/or gives an error. Dropping back to v5.19 and all is good.

Thanks!


Brad

Any update on the UPS Package being broken in v5.20? You can select usb1, but it hangs and/or gives an error. Dropping back to v5.19 and all is good.

Thanks!


Brad

Bump.....

Pavlan how many ports do you have on your bridges?

On the 750G - two corresponding vlan interfaces. But its not config issue.
I've seen these strange readings in host table on various devices.
Especially mac addresses appearing as 00:00:00:00:00:00.

The NV2 wireless on long distance is more stable now in my opinion giving much more reliable throughput and not dropping or giving poor performance occasionally. I think it may now have more issues on shorter distance links now though as I see our shortest links now having occasional disconnects where they didn't on previous versions but some of out long ones had some minor issues on previous versions are now better.

hello Mr.normis, Mr.uldis.

my ip address is 192.168.2.7 connected to RB750GL(ros version 5.20) through vlan, i'm trying to block icmp from other src ip address, so i have rule the rule above is no interface defined, and then i'm trying to ping mikrotik.com from ip address 192.168.2.7, result is RTO. but when i set in-interface=vlan-local icmp is Reply.

i'm trying this problem too using my RB433AH(ros version 5.20) but client not through vlan interface (through ether1), with same rule i ping mikrotik.com and result is Reply.

why that rule won't works without vlan interface defined?

thank you in advance Sir..

RB 433UAH - 100% CPU Load - PPP - SSTP
Ticket#2012083066000695

Had to replace an RB/493 today.

I had (2) 493AH's in stock, brand new in plastic.

1. I powered the first one up, noticed it had 4.17 on it. Was going to take it to 5.20, but I remember what happened with the GL's, so instead I took it to 5.15 first then > 5.20

2. Dumped the backup file on.

3. Installed/replaced old one at tower.

4. Worked PERFECT for 1 hour.

5. All LAN traffic on bridged ports 2-9 COMPLETELY stop.

6. Rebooted, no fix

7. Downgraded firmware back to 5.15, and routerboard firmware, no fix.

8. Took back to 4.17, no fix

9. Un-brigged all the ports, and assigned the IP block to ONE port, still no talk to the AP on the same subnet!

10. RESET CONFIGURATION, re-build the config from scratch. NO FIX.

11. Went back to office, opened up my 2nd RB/493AH spare, brand new. LED Lights Power up and STAY ON, DOES NOT BOOT!!

12. Replaced with RB/433AH + Switch (since it only has 3 ports, I need all 9) = Worked PERFECT. Running 5.20.


Am I just having bad luck lately with Mikrotik or what!!?!?!?!?

wispwest, once again, please send us serial numbers from the affected units. We tested a lot of RB411GL yesterday, and could not find any problems with upgrade (from stock v5.11 to v5.20). Maybe there is something special about you boards, different batch or something like that.

Hello Folks!

SNMP -> 100% CPU again....

We have had problems now and then with 100% CPU, it has always been traced down to SNMP, disable and enable it and it works for some time.

It hits back to the following devices in our environments only)
RB333 about every 3:rd month
RB450G about every 5:th month
SXT5D every month

This happens as well on ros5.20 as it does in older versions in ros5.x no worse none better.

What can this be ?

RB433UAH 100 % CPU SSTP - does not respond to system reboot

> system reboot
Reboot, yes? [y/N]:
y
system will reboot shortly

Without effect.

Steen I've been facing that same problem some 450's.Do yours sometimes lock up in unison or are they hanging separately ? I wonder if there's a way to write a script to restart the SNMP service on a regular basis to see if that doesn't eliminate the problem ?

Hello Folks!

SNMP -> 100% CPU again....

We have had problems now and then with 100% CPU, it has always been traced down to SNMP, disable and enable it and it works for some time.

It hits back to the following devices in our environments only)
RB333 about every 3:rd month
RB450G about every 5:th month
SXT5D every month

This happens as well on ros5.20 as it does in older versions in ros5.x no worse none better.

What can this be ?

I am also seeing this issue. I thought it was specific to RB750GL and RB750UP but we appear to be seeing it on RB2100 as well. I have a ticket logged Ticket#2012081366000093 but Mikrotik are taking a very long time between responses.

Beware rb450g just upgraded one of the version 4.17 to version 5.20 and stopped responding, this team is at 52km away, on a mountain to raise in 4x4
sorry for my bad English

Same here. v5.19 UPS package works fine, but not in v5.20. Put everything back to v5.19 and UPS is good and working. Need this fixed!

Thanks!

Brad

When i try to add UPS - The following errors: "Couldn`t add New UPS - opening serial port failed: 2 9 (6)"
5.19 - all OK

The same problem to me

On RB 450G too many FCS Errors, Align Errors and ethernet ports limiting 5mbps !!. Back to 5.18 ALL OK.

On RB 450G too many FCS Errors, Align Errors and ethernet ports limiting 5mbps !!. Back to 5.18 ALL OK.

Please write to support@mikrotik.com about this problem with attached v5.20 supout.rif file.

On RB 450G too many FCS Errors, Align Errors and ethernet ports limiting 5mbps !!. Back to 5.18 ALL OK.

Please write to support@mikrotik.com about this problem with attached v5.20 supout.rif file.

It is very risky because of that RB450G I have 150 clients and do not want to risk to put 5.20 again.

RB951-2HnD using 5.20
With quickset in ap/cpe mode using router mode. When i set dns server, after apply , setting of DNS server back to empty. I check over export compact, or in /ip dns show empty. This is bug or i missing something ??

I have RB450G and RB493AH, all this time upgraded without any problem, but this time i try from ROS 5.17 jump to ROS 5.20, problem appear :
1. RB 493AH, Simple Queue dissappear completely, but i solve it with restore last backup file before upgrade, now it's fine.
2. RB450G, PPP menu dissappear, PPP client still functions but when open it to configure, only have General and Traffic Tab, nothing else to input, No Dial Up Tab.
3. RB450G, /system/routerboard menu and /routing menu at winbox dissappear too.

I cannot solve RB450G problem even with restoring last backup before upgrade.

Anybody got same problem?

Hello Folks!

100% CPU again, devices affected SEXTANT and SXT all other devices RB411, RB750, RB450G, RB333, RB600 is not affected.

It was all tracked town to SNMP, I have now disabled it (not good monitoring is affected) to see if it helps.

This isssue has been around for a long time now, I fist observed it at this level with reboots every third week or at least one time per month, when we started to use SXT and SEXTANT devicess

Is there any solution ?

Has anyone have issue in ROS 5.20 configuration PCC like in wiki. I mean if we create pppoe-client use usermanager or not and in mangle PCC after disconect pppoe client "red mark" like my picture. Test with RB1200 & RB450

After test why my 450G got problem when upgraded from ROS 5.17 to ROS 5.20, i'm successfully fix it, by downgrade to ROS 5.17 (the problem still same, missing PPP) then upgrade to ROS 5.18 then ROS 5.19 then upgrade firm from 2.39 to 2.41 then upgrade to ROS 5.20 then load my last stable configuration backup while i was ROS 5.17. It's done perfectly no problem. My PPP appear again.

My conclusion : Next time i will upgrade ROS by sequential version, not jump over.

Same here. APC Smart-UPS 620I on RS232 in RB450G

Same here. v5.19 UPS package works fine, but not in v5.20. Put everything back to v5.19 and UPS is good and working. Need this fixed!

Thanks!

Brad

When i try to add UPS - The following errors: "Couldn`t add New UPS - opening serial port failed: 2 9 (6)"
5.19 - all OK

I update one 493AHG, now I can not add an APC ups, please someone could send me the link where you can download 5.19
Thanks for your help

UPS doesn't work with 5.20.
With 5.19 it works very well.
Is there something wrong?

"Couldn't change UPS <ups1> - port's channel is already in use (6)"

The same configuration but with RouterOS 5.19 works fine.

Can you help me?

Thanks.

Same here. APC Smart-UPS 620I on RS232 in RB450G

Same here. v5.19 UPS package works fine, but not in v5.20. Put everything back to v5.19 and UPS is good and working. Need this fixed!

Thanks!

Brad

When i try to add UPS - The following errors: "Couldn`t add New UPS - opening serial port failed: 2 9 (6)"
5.19 - all OK

The same error here, with 5.19 all works fine.
I need help.

Thanks.

uptime 23d RB433UAH (power 24V 2A) error DHCP pool - this is only used pool1 for all IP (default pool2+pool1) + does not load from the table Leases IP > reboot OK

Compared to 5.16 and 5.17 cpu load has doubled with 5.20.
RB411AH wirelss NV2 30 Clients going up from 20-40% to 50 and more even by 4Mbit throughput
And ehternet port flapping is back.

has anyone the same expierence?

Compared to 5.16 and 5.17 cpu load has doubled with 5.20.
RB411AH wirelss NV2 30 Clients going up from 20-40% to 50 and more even by 4Mbit throughput
And ehternet port flapping is back.

has anyone the same expierence?

Many correctly. Example - RB450G with a Sextant-5HND ROS 5.20 . There are many errors FCS, Align error and ethernet flaping

I'm also having problems with the UPS (APC SmartUPS RT 3000). Has been working without problems before 5.20.




Is there any news on this issue?

UPS doesnt work in 5.20, mistake of one of the best mikrotik programators....
You must wait till next release....

It's probably typo in source code. Small error - Why mikrotik not issued the new package? Why do we have to wait for the new full ROS release?

UPS doesnt work in 5.20, mistake of one of the best mikrotik programators....
You must wait till next release....

Did you know what happend to my rb450g after one year awesome performace in hotspot 5.14 , after upgraded to 5.20 then reboot ! Boom ,no dhcp, no static dhcp leases , no 100 hotspot users . Happend like someone deleted them. What a catastrofe .....

Did you know what happend to my rb450g after one year awesome performace in hotspot 5.14 , after upgraded to 5.20 then reboot ! Boom ,no dhcp, no static dhcp leases , no 100 hotspot users . Happend like someone deleted them. What a catastrofe .....

I've been running my RB450G for weeks with 5.18 and upgraded to 5.20 ... after second reboot all graphing data was lost ... so I guess there is something wrong with 5.20 firmware - it deletes data.

I have here some unregular shutdowns on my rb450 devices...

I seem to be having some problems with the POE controller 2.0 firmware. On a 750UP, I upgraded to 5.21, and then manually upgraded the POE controller to 2.0. I'm only using POE out on ether5. Now, whenever ANYTHING is plugged into ether3, ether5 POE cycles power on and off, rebooting the connected Groove every 10 seconds or so. This happens both when ether5 POE is set to auto, or on. Ether3 POE is set to off.

Unfortunately, I updated the POE controller on 6 750UPs before I noticed the problem. They all show this behavior. They did not show this behavior with the 1.1 POE firmware.