Ad blocker detected: Our website is made possible by displaying online advertisements to our visitors. Please consider supporting us by disabling your ad blocker on our website.
Apart from the strange mixing of the ds sequence, eveything seems to be ok.
So I had a look at the snmp get in the log. They seem to be alright again. For RxWords they read
Now I replayed the rrdtool update statements grepped from the cacti.log. With rrdtool xport I dumped one single DS swFCPortRxWords. Everything was ok.
Looking and the brocade xml data template, I saw that nearly all DS had a min of 0 and a max of 100. So I defined another rrd
with those min:max definitions. The rrdtool update with same values was ok, again. But now the dump showed all "NaN"!
So I thing this may be the problem cause. Will try to "rrdtool tune" my productive rrds on monday.
I tried to change the 'data queries' to get the names of the Ports in the title, but I'm not shure what the right values are.
I tried 'query_ifDescription' but that looks like it doesn't work.
Can you help me to get the names? In the 'new graphs' window there you can see the Discription of the port. But how do I get this name in the port title?
for all ds. Here's the result (see graph below, the blue areas!).
I checked all Ports on all switches: no more NaN's. So everything seems to be fine. Now will check the Data Templates. Only swFCPortTxWords has no maximum defined, all other ds have a max value of 100 (which is the default, if you create a Data Template DS).
Thank you again for your templates.
Of course, you have to replace fc2_967.rrd with the name of your own rrd file (to be found in <path_cacti>/rra). You have to repeat this for each rrd file.
Then you have to modify the data template Brocade RC DataTemplate. Each data source item (e.g. swFCPortRxBadEofs should have the value 0 as Maximum Value. This ensures, that each newly defined Switch will have correct Maximum Values.
More info: man rrdtune
Thank for the solution! Work fine now. Do you have an idea what is the relationship bettween the SNMP metric (RxWords/TxWords, RxFrames, TxFrames) and the PortPerfShow ? I have a port that run steady at over 15MB/s and the graph only show average of RxWords 4.47MB/s. and 0 for TxWords and few Kbs for TX/RX Frames.
Anyway. I've upgraded the templates with the max value '0'.
I'm about to write indexed snmp-queries for fan and temp sensors later on. Or if someone wants to contribute with these would be nice.
It seems that the code within the switch-software have pre-defined sensor counters from factory. If the sensors isn't present it'll return a negative value. As I dont know how to sort those sensors out from a basic indexed snmp query I'm afraid i have to write a perl script that filter these sensors out.
After reading few RFC on FC protocol I found that the FC Word is a 4 character of 8 bit so 32bit but, the FC protocol encapsulate those character into a 10bit format so on the pipe, a 32bit data is 40bit.
Now based on a discussion with Brocade, the portpershow show the amount of data non encapsulated. After a test, the SNMP formula:
RxWOrd + TxWord * 4 give the same results as PortPerfShow.
So I did a new CDEF called FCWordSize with this formula:
item #1 : Special Data Source: CURRENT_DATA_SOURCE
item #2 : Custom String: 4
item #3 : Operator: *
Both templates import into 0.8.6d fine, but when I go to add the device I get this error when I click save:
Warning: Variable passed to each() is not an array or object in /var/apache/htdocs/cacti/lib/data_query.php on line 490
Warning: Cannot modify header information - headers already sent by (output started at /var/apache/htdocs/cacti/lib/data_query.php:490) in /var/apache/htdocs/cacti/host.php on line 121
BUT, I can click the back button and go to Devices list and the device is there listed fine.
I also do get:
Data Query [Brocade FC Switch] Reload Associated Query
Error in data query.
When I attempt to create graphs for the hosts.
Any help would be great. These are Silkworm 2400 switches.
