February 2010 - Xymon - Xymon mailman web

ifstat.mac.rrd question
by rmcgraw＠purdue.edu 18 Feb '10

18 Feb '10

I am getting this error message in my rrd-data.log about every minute. I looked in the archive about this and only found one entry that did not provide much help. Also in the trends in the network traffic graph I see an interface named "mac". 2010-02-15 08:41:35 RRD error updating /local/xymon/data/rrd/gauss.math.purdue.edu/ifstat.mac.rrd from 128.210.3.205: /local/xymon/data/rrd/gauss.math.purdue.edu/ifstat.mac.rrd: illegal attempt to update using time 1266241295 when last update time is 1266241295 (minimum one second step) QUESTION: 1) what is the "mac" interface? 2) what is causing the above error and how can I prevent this? Thanks Robert _____________________________________________________________________ Robert P. McGraw, Jr. Manager, Computer System EMAIL: rmcgraw at purdue.edu Purdue University ROOM: MATH-807 Department of Mathematics PHONE: (765) 494-6055 150 N. University Street West Lafayette, IN 47907-2067

3 3

Hobbit: MS cluster aware?
by christopher_bodnar＠glic.com 18 Feb '10

18 Feb '10

Can Hobbit monitor a Microsoft cluster? If so is there any documentation on how to do this? Thank you, Chris Bodnar, MCSE Systems Engineer Distributed Systems Service Delivery - Intel Services Guardian Life Insurance Company of America Email: christopher_bodnar at glic.com Phone: 610-807-6459 Fax: 610-807-6003 ----------------------------------------- This message, and any attachments to it, may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient, you are notified that any use, dissemination, distribution, copying, or communication of this message is strictly prohibited. If you have received this message in error, please notify the sender immediately by return e-mail and delete the message and any attachments. Thank you.

2 1

Cannot create rrd
by olivier＠audry.fr 18 Feb '10

18 Feb '10

hello I have a custom graph who is working well for a host. I want to use it for another but I have this error 2010-02-18 16:08:32 RRD error creating /data/hobbit/data/rrd/*******/gc_event.rrd: min must be less than max in DS definition and the definition is : 2010-02-18 16:08:32 hobbitd_rrd: Got message 883323 @@status#883323|1266505712.239397|84.14.102.145||lmwbs-ws05|gc_event|1266507512|green||green|1266504192|0||0||1266505416 2010-02-18 16:08:32 startpos 3917999, fillpos 3917999, endpos -1 2010-02-18 16:08:32 Creating rrd /data/hobbit/data/rrd/lmwbs-ws05/gc_event.rrd 2010-02-18 16:08:32 RRD create param 00: 'rrdcreate' 2010-02-18 16:08:32 RRD create param 01: '/data/hobbit/data/rrd/lmwbs-ws05/gc_event.rrd' 2010-02-18 16:08:32 RRD create param 02: '-s' 2010-02-18 16:08:32 RRD create param 03: '300' 2010-02-18 16:08:32 RRD create param 04: 'DS:YGC:DERIVE:600:U:U' 2010-02-18 16:08:32 RRD create param 05: 'DS:FGC:DERIVE:600:U:U' 2010-02-18 16:08:32 RRD create param 06: 'RRA:AVERAGE:0.5:1:576' 2010-02-18 16:08:32 RRD create param 07: 'RRA:AVERAGE:0.5:6:576' 2010-02-18 16:08:32 RRD create param 08: 'RRA:AVERAGE:0.5:24:576' 2010-02-18 16:08:32 RRD create param 09: 'RRA:AVERAGE:0.5:288:576' 2010-02-18 16:08:32 RRD error creating /data/hobbit/data/rrd/lmwbs-ws05/gc_event.rrd: min must be less than max in DS definition thx for your help oau

1 1

strange rrd data
by olivier＠audry.fr 18 Feb '10

18 Feb '10

hello since yesterday we have some strange data in our rrd. Data totally out of range (e+150). I have enable the debug mode to see what information hobbit get and send. Someone has already have this kind of trouble ? thx for your help hobbit version 4.2.0 rrd version 1.2.23 oau

2 2

Here are possible patches affecting hobbitfetch
by cade.robinson＠gmail.com 18 Feb '10

18 Feb '10

I have been having some issues with hobbitfetch and have modified a few files to fix them. At least my definition of fixed in that hobbitfetch now works - but the way they are fixed may not be right. include/libbbgen.h Change IP_ADDR_STRLEN from 16 to 22. The reason is that I was getting reports from IP:port and that is 22 chars with the trailing NULL. At 16 I was getting multi-source reports and the the host shown in "Status message received from" was truncated otherwise. hobbitd/hobbitd.c Again related to the above - took out he %16s because my host could be 22 chars. hobbitd/hobbitfetch.c Lowered res from 100 to 25 - same as above so it could really be 22. Not much of a change but save a bit of mem. Change the read to read two less bytes than sizeof(buf) and set buf[n-1] to NULL. This is in case 8192 bytes are sent/read. Since buf is 8192 elements long the last element is 8191 not 8192 so buf[n]=\0; when 8192 bytes are read causes a segfault. I think the same should be done in msgcache.c as well but I haven't seen anything over 100 bytes be sent to msgcache. Here are the patches: Index: include/libbbgen.h =================================================================== --- include/libbbgen.h (revision 6223) +++ include/libbbgen.h (working copy) @@ -27,7 +27,7 @@ #define STRBUF(buf) (buf->s) #define STRBUFLEN(buf) (buf->used) -#define IP_ADDR_STRLEN 16 +#define IP_ADDR_STRLEN 22 #include "version.h" #include "config.h" Index: hobbitd/hobbitd.c =================================================================== --- hobbitd/hobbitd.c (revision 6223) +++ hobbitd/hobbitd.c (working copy) @@ -2789,7 +2789,7 @@ /* Pick out the real sender of this message */ msgfrom = strstr(currmsg, "\nStatus message received from "); if (msgfrom) { - sscanf(msgfrom, "\nStatus message received from %16s\n", sender); + sscanf(msgfrom, "\nStatus message received from %s\n", sender); *msgfrom = '\0'; } @@ -2866,7 +2866,7 @@ else if (strncmp(msg->buf, "status", 6) == 0) { msgfrom = strstr(msg->buf, "\nStatus message received from "); if (msgfrom) { - sscanf(msgfrom, "\nStatus message received from %16s\n", sender); + sscanf(msgfrom, "\nStatus message received from %s\n", sender); *msgfrom = '\0'; } @@ -2907,7 +2907,7 @@ msgfrom = strstr(msg->buf, "\nStatus message received from "); if (msgfrom) { - sscanf(msgfrom, "\nStatus message received from %16s\n", sender); + sscanf(msgfrom, "\nStatus message received from %s\n", sender); *msgfrom = '\0'; } @@ -3668,7 +3668,7 @@ if (msgfrom) { char *ipline = strstr(msgfrom, "\nClientIP:"); if (ipline) { - sscanf(ipline, "\nClientIP:%16s\n", sender); + sscanf(ipline, "\nClientIP:%s\n", sender); } } Index: hobbitd/hobbitfetch.c =================================================================== --- hobbitd/hobbitfetch.c (revision 6223) +++ hobbitd/hobbitfetch.c (working copy) @@ -100,7 +100,7 @@ char *addrstring(struct sockaddr_in *addr) { - static char res[100]; + static char res[25]; sprintf(res, "%s:%d", inet_ntoa(addr->sin_addr), ntohs(addr->sin_port)); return res; @@ -346,7 +346,7 @@ char buf[8192]; /* Read data from a peer connection (client or server) */ - n = read(conn->sockfd, buf, sizeof(buf)); + n = read(conn->sockfd, buf, sizeof(buf)-2); if (n == -1) { /* Read failure */ time_t now = gettimer(); @@ -361,7 +361,7 @@ /* Save the data */ dbgprintf("Got %d bytes of data from %s (req %lu)\n", n, addrstring(&conn->caddr), conn->seq); - buf[n] = '\0'; + buf[n+1] = '\0'; addtobuffer(conn->msgbuf, buf); } else if (n == 0) {

2 1

SPLITNCV
by steinarms＠gmail.com 17 Feb '10

17 Feb '10

Hi all, I have create a test that I call queue_count and it monitors the QueueDepth of JmsQueue's on Weblogic My output is like so : OnlineQueue : 5 TimeQueue : 0 ReloadQueue : 1 Queue names are not always the same between hosts so I'd rather not have the queue name as a DS in the rrd. I put my check in the hobbitserver.cfg TEST2RRD includes "queue_count=ncv" SPLITNCV_queue_count=*:GAUAGE:600:0:U" ( also tried *:GAUAGE ) I always get a similar error in rrd-satus.log 2010-02-16 09:32:05 hobbitd_rrd: Got message 1604 @@status#1604/calkopas17|1266312725.234658|10.10.10.10:198 ||calkopas17|queue_count|1266314525|green||green|1266271210|0||0||1266312669|linux|external_servers/Test 2010-02-16 09:32:05 startpos 392396, fillpos 392396, endpos -1 2010-02-16 09:32:05 Creating rrd /var/lib/xymon/rrd/calkopas17/queue_count,ReloadQueue.rrd 2010-02-16 09:32:05 RRD create param 00: 'rrdcreate' 2010-02-16 09:32:05 RRD create param 01: '/var/lib/xymon/rrd/calkopas17/queue_count,ReloadQueue.rrd' 2010-02-16 09:32:05 RRD create param 02: '-s' 2010-02-16 09:32:05 RRD create param 03: '300' 2010-02-16 09:32:05 RRD create param 04: 'DS:lambda:DERIVE:600:U:U' 2010-02-16 09:32:05 RRD create param 05: 'DS:lambda:DERIVE:600:U:U' 2010-02-16 09:32:05 RRD create param 06: 'RRA:AVERAGE:0.5:1:576' 2010-02-16 09:32:05 RRD create param 07: 'RRA:AVERAGE:0.5:6:576' 2010-02-16 09:32:05 RRD create param 08: 'RRA:AVERAGE:0.5:24:576' 2010-02-16 09:32:05 RRD create param 09: 'RRA:AVERAGE:0.5:288:576' 2010-02-16 09:32:05 RRD error creating /var/lib/xymon/rrd/testserver1/queue_count,ReloadQueue.rrd: Duplicate DS name: lambda So it looks like it's not picking up the "GAUGE" and it's always trying to squeeze two results into one rrd where I want a rrd per JmsQueue I'm using xymon 4.3.0-beta2 Do you spot some obvious errors in my config ? Best Regards, Steinar M.

3 5

purple hosts - better fix
by cade.robinson＠gmail.com 17 Feb '10

17 Feb '10

I think this is a better fix for purple hosts. I was thinking about why there was the while loop and I am guessing there could be a duration like: 1d4h10m So with that I left the while loop in and at the end of the block where startofval is set if *p I changed the test since space counts as a valid char for *p. This is in the 4.4.0 branch so the line numbers may not match for other source versions. Index: lib/timefunc.c =================================================================== --- lib/timefunc.c (revision 6223) +++ lib/timefunc.c (working copy) @@ -475,7 +475,7 @@ } result += oneval; - startofval = ((*p) ? p+1 : NULL); + startofval = ((*p) ? ((isspace((int)*p)) ? NULL : p+1) : NULL ); } return result;

1 0

Problems with dbcheck.pl
by linuxmasterjedi＠free.fr 17 Feb '10

17 Feb '10

Hello, I'm writing today for issues regarding the Hobbit extension "dbcheck.pl". Since i've updated the script to 1.09, it has a strange behavior. First of all, my MySQL server is stopped around 23h30 for offline backup. When Hobbit launch "dbcheck.pl fast" at this time, I can see in the logs : "Tue Feb 16 23:35:35 2010:ERROR on mysql: Can't connect to MySQL server on 'mysql.mydomain.tld.local' (111)" That's quite normal. The strange behavior is the column "MySQLPerf" "TblSpace" appears and become purple. Do you know why ? TblSpace is more Oracle related informations, now MySQL. Secondly, I have a bunch of warning in the logs : Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Tue Feb 16 23:35:35 2010:ERROR on mysql: Can't connect to MySQL server on 'mysql.latecoere.gl.local' (111) Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Undefined subroutine &Hobbit_fd_lib::get_numeric_value called at /home/users/hobbit/application/server/ext/dbcheck.pl line 361. Here is the part of the dbcheck.pl involve : 360 my $alertthreshold; 361 ($warnthreshold,$alertthreshold)=Hobbit_fd_lib::get_numeric_value($hostname,"threadswarnalert",500,1000); 362 if($thr>$alertthreshold) { Is there a way to fix this issue ? Thanks by advance,

1 0

TS - Disable until OK
by d.tom.schmitt＠L-3com.com 16 Feb '10

16 Feb '10

Is there a way to disable from a command line until OK. I can run the bb command with disable but I cannot find an until OK option like on the web page for disable. I believe that this takes a device out of the SLAs. I have a request from the network team to be able to suspend maintenance in certain areas. There is no set time frame or it could be unscheduled maintenance. We have almost 1,000 devices and I need to create scripts say for all the DMZs on a firewall, or all the devices on a switch. I can then have them run a script from a web page to stop all the alerting during the outage. Thanks, Tom Schmitt Senior IT Staff - R&D L-3 Communication Systems West 640 North 2200 West P.O. Box 16850 Salt Lake City, UT 84116 Phone (801) 594-3030 Cell (801) 231-7230 eFax (413) 480-6873 D.Tom.Schmitt at L-3Com.com \\\\||//// \ ~ ~ / | @ @ | --oOo---(_)---oOo--

2 3

purple hosts - SOLVED.
by cade.robinson＠gmail.com 16 Feb '10

16 Feb '10

I think I finally have the fix to purple hosts when the host starts with number! The REAL culprit is lib/timefunc.c in durationvalue function. --- lib/timefunc.c (revision 6223) +++ lib/timefunc.c (working copy) @@ -456,7 +456,7 @@ startofval = dur; - while (startofval && (isdigit((int)*startofval))) { + if (startofval && (isdigit((int)*startofval))) { char *p; char modifier; int oneval = 0; Can anyone tell me why the parse code starting with that "while" would need to execute more than once? >From what I can see the messages passed in are either a duration only or a message that is: validity hostname.test color blah blah So it should only have to parse the string once - correct?

1 0