I own a RB493G board which is very underutilized (only routing 3 users). I have a problem with it, I get intermittent packet loss on ethernet ports at random intervals (measured in tens of seconds). When surfing the web is a bit annoying since some pages get stalled but voip is horrible.
I simply test by pinging routerboard (normal pings and flood pings), and suddenly some packets are not replied. For example
64 bytes from 192.168.55.1: icmp_seq=38 ttl=63 time=3.646 ms
64 bytes from 192.168.55.1: icmp_seq=39 ttl=63 time=8.572 ms
64 bytes from 192.168.55.1: icmp_seq=40 ttl=63 time=3.912 ms
Request timeout for icmp_seq 41
Request timeout for icmp_seq 42
64 bytes from 192.168.55.1: icmp_seq=43 ttl=63 time=3.447 ms
64 bytes from 192.168.55.1: icmp_seq=44 ttl=63 time=3.244 ms
64 bytes from 192.168.55.1: icmp_seq=45 ttl=63 time=3.646 ms
64 bytes from 192.168.55.1: icmp_seq=46 ttl=63 time=3.234 ms
64 bytes from 192.168.55.1: icmp_seq=47 ttl=63 time=3.077 ms
I am using RouterOS 5.19, have downgraded to 5.17, netinstalled 5.19, reseted to defaults and I get the same results. In fact, just after factory reset if I only configure an IP address to ping I get the same results. Any ideas? Faulty board?
The board is grounded to the same ground of other connected equipment and the routerboard is about 4 months old. I have noticed this problem some weeks ago, but I can not assure if this happened when was new.
Today I have powered off the board for about 1 hour (disconnected from power) and then I have backed up the configuration, reseted to defaults and then loaded the previously saved backup. After 6 hours pinging about 680 packets/s no packets have been lost! What is happening?
After more extensive testing I get no more packet lost using ping flood, but I still get random high latencies. I have been investigating more the “randomness” and I see it occurs every 70 seconds (more or less). Sometimes latencies are so high packets are lost, other times simply I get latencies higher than 500 ms.
I am using firmware version 2.41 and ROS 5.19. No metarouter running.
Thank you for you reply.
I have disabled watchdog and the same result. Reboot and the same. Now for the tests I am using ether7, but the same happens with other ports. I have not configured anything under the switch options, they are in factory defaults.
Any other idea? I have been looking at scheduled actions, and I only see a DNS script run every 5 min. I have also tried disabling netwatch (which is also using 5 min). Any other periodic task to look at?
Have you tried a system reset and only have one PC connected with DHCP enabled?
No other connections, only one PC.
Try the ping again, is it still the same problem?
Sill the same issue. Tested on all ports. All ports are 100mbps auto-negotiation enabled.
I have been looking a bit more, and I see voltage in system health always is around 22.6 volts. I am using a 24w (24v x 1A) power supply. There may be a problem with the power supply?
I have only 3 ethernet ports connected, one mini-pci wlan card and a micro-sd. Should I change the power supply for a more powerful one? According to specs of 493G power consumption is ~3W without extension cards, maximum – 25W (18W output to extension cards). I don’t think the wlan card is consuming so much, most of the time is idle.
Tried an other PC?
Tried without the SD and the mini-pci card?
And make shure that you don’t have anything else connected, even if the interface are disabled.
This is common problem with 493G - we noticed as soon as we started upgrading 493 with 493G (routing masthead 411 and 711 base station units) - problem solved by replacing with RB1100AH.
Having the same issue.
-RB493G properly grounded with original Mikrotik case.
-Powered through POE 24V 1A and PowerJack 12V 5A at the same time.
-Running 5.20 OS with 2.41 firmware
-Previous 1100AH worked flawlessly for 9 months in its place.
It is NOT a temp issue, it is NOT a powersupply issue. Wireless traffic works normal, just the ethernet is having this problem. It is most likely a switch chip issue.
Ive got a large MT network (RB493G) using EoIP tunnels going from each site to a main PPPoE GW (RB1100AHx2),
i had similar issues with dropping packets on the ether ports, one fix that cured these 10 boxes was i downgraded it to v4.17 this stableised the entire network which has been running happily for about 10 months,
the MT consulant who we liaze with thought it could have been the chipset in the newer versions (around the 5.14-5.18 at the time) has a different chipset firmware. (i could be recalling the conversation wrong LOL!)
try this and see how you get on,
im currently trialing v5.20 on a RB493G with a constant ping to see if any packet loss or “port Flapping” occurs
Kev
Current network consists of 10xRB493G’s, 2xRB2011L-IN and 1xRB1100AHx2 (with a RB1100AH backup) running PPPoE, OSPF and EoIP tunnels