*) fixed static object crash
*) changed icon
*) some minor fixes
Good day! Still not updated charts through a Web page in one session!
Any chance to be able to specify a different RouterOS port for monitoring Mikrotik devices?
4b3 observations - Backup your dude folder before upgrading. Upgrade was smooth as usual.
As noted FIXED - If you click “add static” the dude no longer crashes.
Not Fixed - The remove resolved button on the outages tab of services doesn’t remove resolved outages.
Other Features
- Exporting a backup stops the polling process, any probes that time out while the export is being done will appear down soon after the export has finished.
2)Select a device then click the copy button all the history will be erased and graphs will be gone.
3)On the probe panel select a probe then click the copy button, the probe XML will include all the devices it is installed on.
NOTE: Care should be taken to remove the unneeded XML when copying and pasting or sharing probes through XML.
Still nothing to worry about KEEP up the great work!
any update about monitor network , and show usage graph for mikrotik …
Some probes shown as “not available” in device services. New Function probe has a default text Available=1, Value=2+2.
If I delete the old probe and makes a new similar one it will work, but it gives me a lot to do… Fix in a hurry??
Waiting for next release! thanks!
@eivind Don’t delete the old probe. Wait at least 6 minutes to see if the old probe starts working.
“not available” can mean that the previous SNMP read did not work that would cause a “false positive”. If the previous read did not work then the probe will not work for 5 minutes or more specifically the amount of time that is set in Negative Cache time of the function that grabs the OID. The default of which is 300 seconds.
You can modify the Negative Cache time but it takes a bit of work by modifying every function. If you do make it lower your false positives will not be 5 minutes long.
I hope that makes sense,
Lebowski
Thanks Lebowski, I know. Some faulty probes did come up after five minutes. I did a test by making one new identical probe with a different name. It did work, but the old ones who has been OK since beta 2 arrived still don’t work after 24 hurs. I’m testing if it is available with “function() >0” and have other probes who is working fine with exactly the same test on a close to identical function. I tried to set “” in the test but it still did’nt work. There might be a communication problem between functions and probes. As one said in this forum; It’s a dangerous life to always use latest version, but since the Dude is free of charge I’m not complaining, just telling 
“function() >=0” In version 4b3 the “Available” line is less forgiving.
If you take the probe that doesn’t work and add one to the value that is being read the probe comes up but it will never appear down. I have solved this for my self by following gsandul’s example of a function that returns “False” or a value and a probe with two if statements.
Function: if(array_size(oid_column(“1.3.6.1.4.1.9.2.1.57”, 10 ,29)), oid(“1.3.6.1.4.1.9.2.1.57.0”, 10, 29)+1 ,“False”)
If the oid is available read the value and add one to it otherwise return false.
Probe: if(Cisco_CPU_a()<>“False”,if(Cisco_CPU_a() -1< 70, “”, concatenate("Warning: high CPU = ", Cisco_CPU_a(), “%”)), “Cisco Device down”)
If the value is not “False” subtract 1 from it make sure the value is less than 70 or complain about high CPU, If the value is False return “Cisco Device Down”.
not fixed bugs:
New Network map only says ‘loading’
Bug - logs reseted to default after Windows reboot
Solved it the easiest way! Migrating this formula into gsandul’s solution would give me a real challenge:
Uncorrectables x 100 / (Unerroreds + Correcteds + Uncorrectables)
(oid(“1.3.6.1.2.1.10.127.1.1.4.1.4.3”)*100)/
(oid(“1.3.6.1.2.1.10.127.1.1.4.1.2.3”)+(oid(“1.3.6.1.2.1.10.127.1.1.4.1.3.3”))+(oid(“1.3.6.1.2.1.10.127.1.1.4.1.4.3”)))+0.001
Maybe not the smartest way of doing it, but it does the job.
Because Dude don’t graph under errors I have chosen to make a separate probe for “Data Correction Too High” , and using another single oid probe for down statement for the devices.
I’m trying as hard as I can to keep my networking knowledge at a minimum, and give myself more time to run my business.
Thank You for Your inputs
is it released for PPC (RB1100)? - seems like only beta1 files are posted on the site.
If not, then is there an ETA (can you let us know what the issue is?)
thanks,
Hi! I have some backup problems, after updating my dude from v3.6 to 4beta, 4beta2 and 4beta3. when i do backup (using dude client) dude server terminate my connection, and all connections from other users. Dude client runs on windows vista and windows xp…
@caterpillar, The backup is taking too long to export, lower your raw keep time to make sure dude.db doesn’t get bigger than 2GB. Use copy to make a backup of your dude folder for now. Hopefully they fix both the 2GB limit and the issue with IO affecting polling soon.
Still no elbows in links?
I use the “network” element without any data (creates a small yellow cloud that you can use as “elbow” (if you mean this…)
I am also having the “Not Available” services sometimes now… wonder what got changed, there…
Andreas
Lebowski, I would like to discuss using (partial) snmpwalks in functions.. as I am having severe CPU issues here.
Can we fork a new thread to discuss this, please?
My approach for a sanity-check in a function is:
never mind that, it doesn not work
Andreas
Hi …
I went from 4b2 to 4b3 some days ago, no issues to upgrade.
Spent some weeks playing with ROS 5RC11, 5RC12 and now settle on 5.0.
My system is small, e.g. a 433UAH, a pair of 411ARs, 3 old WRAP X86, all loaded with ROS5.0 now. Around 70 CPEs (NS2 Loco & Bullet 2). There is 3 XR9 for NLOS path.
-
Any card that run as client-bridge (R52Hn, XR9, R52n), after this double upgrade to 4B3 x ROS 5.0 missed the speed/signals levels at the labels. On the other side, if I create a link between any UBNT client and AP still showing signals levels (tx/rx), connection speed & etc. At the MT devices, does not matter if mastering is snmp or routeros. Just data speed shows up.
-
Torch avialable at mouse over link lables + right click does not run.
-
I’m using Dude to collect & graph some data volume info, as measured by “bytes” field on queue tree. Those queues are common low priority (priority=8) channels I set to those traffic which customers does not realize about, like Windows Update, several anti-virus update, adobe update, etc … and Youtube/4shared. Each uses pcq and a max limit so customers share those channels. I expected a monthly basis graph, reseting all counters after 30 days by myself - but - most are reseting when the volume reaches 4.1 … 4.3 GBytes. I guess this is not Dude related but ROS. Anyway.
PS: not Dude but 5.0 winbox … torch on simple queues (mouse over a simple queue line) is not running too, the source address filter field on torch is not being filled with simple queue target address, as used to be on ros 4.17.
Regards
So the 2GB DB size limit hasn’t been fixed in this release? Has anyone confirmed this and have Mikrotik acknowledged this bug? This is the only thing preventing us upgrading from 3.6.