From: R.j. W. <rj...@at...> - 2002-10-22 00:14:51
|
where the link to download your app? looks nice=20 r.j. i would like to try it out please |
From: Bhavna R. <bha...@ho...> - 2003-04-16 09:36:46
|
hello, I have installed jffnms and all components on W2K and it seems to be = working except that it is not giving me any performance graph saying = "There is No Information Yet, probably the poller has not created this = interface file(s)". I don't understand where the problem is. And also i = cannot monitor a linux machine which is on the same lan and a netopia = router. I would be very gratefull if you could please give me some solutions = to my problems. Best regards. Thank you, Miss B. Ramkhelawon (Preparing for sys admin in local firm) |
From: Anders K. <and...@so...> - 2003-04-16 09:51:27
|
=2D----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, I have no firsthand experience in running jffnms on W2K, I do know that you= =20 need to run somekind of "cron" job, maybe windows scheduler could do the jo= b=20 ? Anyway, here's what is needed, */1 means every minute, */5 every 5 minut an= d=20 so on. */1 * * * * apache php -q /opt/jffnms/engine/consolidate.php >/dev/null 2>&1 */5 * * * * apache php -q /opt/jffnms/engine/poller.php >/dev/null 2>&1 */30 * * * * apache php -q /opt/jffnms/engine/rrd_analizer.php >/dev/null 2= >&1 */30 * * * * apache php -q /opt/jffnms/engine/autodiscovery_interfaces.php = >=20 /dev/null 2>&1 02 4 * * * apache php -q /opt/jffnms/engine/tftp_get_host_config.php=20 >/dev/null 2>&1 02 5 * * * apache php -q /opt/jffnms/engine/cleanup_raw_tables.php >/dev/nu= ll=20 2>&1 You must configure your W2K machine to run these scripts, otherwise, jffnms= =20 won't work. Regards Anders Karlsson onsdagen den 16 april 2003 11.35 skrev Bhavna Ramkhelawon: > hello, > I have installed jffnms and all components on W2K and it seems to be > working except that it is not giving me any performance graph saying=20 > "There is No Information Yet, probably the poller has not created this > interface file(s)". I don't understand where the problem is. And also i > cannot monitor a linux machine which is on the same lan and a netopia > router. I would be very gratefull if you could please give me some > solutions to my problems. Best regards. Thank you, > Miss B. Ramkhelawon > (Preparing for sys admin in local firm) =2D --=20 SongNetworks and...@so... Computers are like air conditioners. They stop working when you open Windows. Disclaimer: the above is the author's personal opinion and is not the opini= on=20 or policy of his employer or of the little green men that have been followi= ng=20 him all day. =2D----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (GNU/Linux) iD8DBQE+nSbztC1yigtSyRkRAmgJAJ0XRrTptCvFaEv1DVQzYuzaB0XOvACfam9R jlTJNR6J8XZx2tQMzLBDcPk=3D =3Du8J/ =2D----END PGP SIGNATURE----- |
From: Javier S. <ja...@sz...> - 2003-04-16 11:54:52
|
Hi, Which JFFNMS version do you have? Did you follow the INSTALL.win32.txt file ? Did you configured the Scheduled Tasks? Where did you install jffnms? Can you send us the contents of /conf/config.ini ? Can you send us the output of /admin/adm/test.php ? Can you send us the output of "c:\php\php.exe -q c:\jffnms\engine\poller.php"? Javier On Wednesday 16 April 2003 06:35, Bhavna Ramkhelawon wrote: > hello, > I have installed jffnms and all components on W2K and it seems to be > working except that it is not giving me any performance graph saying > "There is No Information Yet, probably the poller has not created this > interface file(s)". I don't understand where the problem is. And also i > cannot monitor a linux machine which is on the same lan and a netopia > router. I would be very gratefull if you could please give me some > solutions to my problems. Best regards. Thank you, > Miss B. Ramkhelawon > (Preparing for sys admin in local firm) |
From: cristi <cr...@my...> - 2003-06-11 18:12:44
|
confirm 283436 |
From: Aurelien B. <aba...@do...> - 2003-06-24 09:24:58
|
Hello JFFNMS, Can someone explain me why i can't display graphs ? My JFFNMS setup is ok, but i have nothing when trying to display a graph. But i have the .png is present in the correct directory. What will i do ? Rgds, Aurelien |
From: Pierre-Damien G. <pd...@fr...> - 2003-12-04 11:24:51
|
Hi, I'm setting up a cluster (HA and load-balancing) of jffnms. I have one server active which it polls differement element. The home directory of jffnms and the mysql database are located on a disk which are mirrored through tcp/ip (drbd). So my active server has read/write access to this partition and the second server has a read-only access via a samba share. When I do a request on my cluster, there is no problem when it's the active server responding me. But when it's the second, I have the errors : The RRDTool files for Interface ID 11 (from Host ID 3), has not been created by the Poller Process yet. The RRDTool files for Interface ID 10 (from Host ID 3), has not been created by the Poller Process yet. I think it's a problem of rights. Do we need a read/write access to those files ? If yes, why ?. And I see also that the temporary image files are not created correctly on the second server (size = 0), cause of we can't open the rrd file... Thanks. Pierre ------------------------------------------------- This mail sent through IMP: http://horde.org/imp/ |
From: Quinton H. (Q) <Ho...@te...> - 2004-01-20 11:36:34
|
Hi, Has anybody discovered Solaris 6 and HPUX 11 with JFFNMS? I have no issues with Lnux 7.3 or Windows 2000 and up. I am experiencing endless hassles with this. The only instance where I get a response is with HPUX 1i and then also, very limited. Cheers Quintin |
From: Tshering P. <pen...@ho...> - 2004-04-05 05:02:39
|
dear sir, I have tried the other version of MySQL(i.e.,) as u said and it worked for us but we r having a poblem connecting to the server. It says tht we have to enter yourserver/jffnms/ ' to access jffnms . And when i last time mailed to the list , they told me that yourserver is the url of the server running jffnms.I had ealier, while installing apache, given the servername to be the work-group name here(ie., un-bt), because we r not being able to get the desired result using any other name. But unfortunately now we can't access jffnms using this. I am going to mail the list about this and i hope it will help. Also , in the installation guide it says tht, to create a user in windows user manager and to give it FULL CONTROL over the c:jfnms folder. We created the user by going to : control panel->users&password -> andthen created the user and we gave it adminitrator rights also. I thought by doing this we will give it full control over the folder.... am i right..r is there another way of giving full control over a folder... And one last thing while installing MYSQl, it says to put an admin Username and password. Do we put it in my.ini Setup, the user name we created for jffnms. please help, your earlier help has been invaluable for me. I am using: apache 1.3.2 MySQl 4.0.18 PHP 4.3.4 JFFNMS 0.7.6 Windows 2000 thanks, sincerely, Tshering Penjor _________________________________________________________________ MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*. http://join.msn.com/?page=features/virus |
From: Javier S. <ja...@jf...> - 2004-04-06 12:44:09
|
Hi, It seems like you are not running W2K or WXP... But what you say is correct, a user with Administration privileges is enoght. Another way is to look at the permissions of the folder and assign it there. You don't have to touch the my.ini. Just follow the INSTALL.win32 instructions. Javier On Monday 05 April 2004 01:50, Tshering Penjor wrote: > dear sir, > > I have tried the other version of MySQL(i.e.,) as u said and it worked > for us but we r having a poblem connecting to the server. It says tht we > have to enter > yourserver/jffnms/ ' > to access jffnms . And when i last time mailed to the list , they told me > that yourserver is the url of the server running jffnms.I had ealier, while > installing apache, given the servername to be the work-group name here(ie., > un-bt), because we r not being able to get the desired result using any > other name. But unfortunately now we can't access jffnms using this. I am > going to mail the list about this and i hope it will help. > Also , in the installation guide it says tht, to create a user in windows > user manager and to give it FULL CONTROL over the c:jfnms folder. We > created the user by going to : > control panel->users&password -> andthen created the user and we gave it > adminitrator rights also. I thought by doing this we will give it full > control over the folder.... am i right..r is there another way of giving > full control over a folder... > > And one last thing while installing MYSQl, it says to put an admin Username > and password. Do we put it in my.ini Setup, the user name we created for > jffnms. > > please help, your earlier help has been invaluable for me. > > I am using: > apache 1.3.2 > MySQl 4.0.18 > PHP 4.3.4 > JFFNMS 0.7.6 > Windows 2000 > > thanks, > sincerely, > Tshering Penjor > > _________________________________________________________________ > MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*. > http://join.msn.com/?page=features/virus > > > > ------------------------------------------------------- > This SF.Net email is sponsored by: IBM Linux Tutorials > Free Linux tutorial presented by Daniel Robbins, President and CEO of > GenToo technologies. Learn everything from fundamentals to system > administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click > _______________________________________________ > jffnms-users mailing list > jff...@li... > https://lists.sourceforge.net/lists/listinfo/jffnms-users |
From: Steve R. <st...@n2...> - 2004-07-05 21:13:16
|
Am all set up on my solaris box, but I cant seem to get this right, the index.php and start.php come up just fine, but then trying to browse to another level I get the following error, The requested URL /jffnms/admin/menu_frame.php was not found on this server. And form what I gather it should be looking for htdocs/admin/menu_frame.php and not jffnms/admin/menu_frame.php Any ideas |
From: Javier S. <ja...@jf...> - 2004-07-05 21:23:32
|
Hi Steve, It seems like an error in your apache configuration.. Please look at the apache's access_log and error_log Did you use a Virtual Domain to host JFFNMS or a Script directive? Javier Steve Rieger wrote: > Am all set up on my solaris box, but I cant seem to get this right, the > index.php and start.php come up just fine, but then trying to browse to > another level I get the following error, > > > > > > The requested URL /jffnms/admin/menu_frame.php was not found on this server. > > > > And form what I gather it should be looking for > htdocs/admin/menu_frame.php and not jffnms/admin/menu_frame.php > > > > > > Any ideas > -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: Pierre-Damien G. <pd...@fr...> - 2004-09-02 21:57:42
|
Thanks for your reply. But my script returns, for example 872, but the sn= mp function return "872" : [root@relay2 engine]# snmpwalk -v 1 192.168.4.20 public .1.3.6.1.4.1.202= 1.51 enterprises.ucdavis.51.1.1 =3D 1 enterprises.ucdavis.51.2.1 =3D "mail" enterprises.ucdavis.51.3.1 =3D "/usr/local/bin/stat_mail.sh" enterprises.ucdavis.51.100.1 =3D 0 enterprises.ucdavis.51.101.1 =3D "872" enterprises.ucdavis.51.101.2 =3D "21" enterprises.ucdavis.51.101.3 =3D "19" enterprises.ucdavis.51.102.1 =3D 0 I try this : if (is_string($value)) settype ($value, "integer"); But it doesn't work. Do you know how we can remove the quotes ? Pierre _________________ Hi Pierre, Everything is fine. Except that the return value is a string, and it has to be a number. You don't see it in the RRD line because its not a number. So, you have 3 options. - Modify the stat_mail.sh to return an INTEGER without the " " - Create a poller to get this OID and convert it to integer. - Modify the snmp_counter poller to convert the return value to integ= er (remove the quotes) I think the first option is the best. I hope this helps. Javier Pierre-Damien Gatouillat wrote: > Hello everybody, > > I have had a new poller item like that : > ID Description Name (Match RRD Struct DS) Poller C= ommand (file) > Parameters > 10001 Mail Accounting Total mail_delete snmp_counter > .1.3.6.1.4.1.2021.51.101.1 > > The OID is a private one and exec un shell script via snmpd.conf : > exec .1.3.6.1.4.1.2021.51 mail /usr/local/bin/stat_mail.sh > > This OID works good and return a value : #snmpwalk -v 1 localhost publi= c > .1.3.6.1.4.1.2021.51.101.1 : > enterprises.ucdavis.51.101.1 =3D "1171" > > Next add this poller to "Linux/Unix System Info" Interface Types : > ID Description Internal Name Position Inter= face Type Field > Type Show Overwritable Tracked Default Valu= e > 10007 Mails Total mail_total 130 Linux/Unix Sy= stem Info RRDTool > DS Never Type: Gauge Min: 0 Max: 10000 or Use= for Max > > Next add an entry in "Poller/Backend Grouping" in Linux/Unix Host: > 10006 Linux/Unix Host 75 Mail Accounting Total =20 Temporal Buffer > > And a graphe type : > 10001 Mails Linux/Unix System Info mail_stats= =20 500 175 > 0 0 > > And a new php script : mail_stats.inc.php with a copy of the rtt graph,= change > the value rtt by mail_stats. > > When I run php -q poller.php 4 41, the rrd file is created, I can see a= graph, > but after new polling, the rrd is not change, and no value in it (rrdto= ol dump > interface-41-12.rrd) > THe result of the poller : > bash-2.05$ php -q poller.php 4 41 > : H 4 : Poller Start : 14 Items. > : H 4 : I 41 : P 10 : snmp_counter:cpu_nice_ticks(.1.3..1.0): 1= 305 -> > buffer(): 1 (time P:3.19 | 0.34) > : H 4 : I 41 : P 15 : snmp_counter:num_procs(.1.3..6.0): 261 -> > buffer(): 2 (time P:1.4 | 0.12) > : H 4 : I 41 : P 20 : snmp_counter:cpu_user_ticks(.1.3..0.0): 4= 3094 -> > buffer(): 3 (time P:1.12 | 0.12) > : H 4 : I 41 : P 25 : snmp_counter:num_users(.1.3..5.0): 0 -> buffer(): > 4 (time P:0.92 | 0.11) > : H 4 : I 41 : P 30 : snmp_counter:cpu_idle_ticks(.1.3..3.0): 6= 540643 -> > buffer(): 5 (time P:1.08 | 0.11) > : H 4 : I 41 : P 35 : snmp_counter:tcp_established(.1.3..9.0): = 68 -> > buffer(): 6 (time P:0.84 | 0.11) > : H 4 : I 41 : P 40 : snmp_counter:cpu_system_ticks(.1.3..2.0):= 79668 -> > buffer(): 7 (time P:1.08 | 0.11) > : H 4 : I 41 : P 45 : snmp_counter:tcp_active(.1.3..5.0): 46088= -> > buffer(): 8 (time P:0.71 | 0.11) > : H 4 : I 41 : P 50 : snmp_counter:load_average_1(.1.3..3.1): 0= .06 -> > buffer(): 9 (time P:0.84 | 0.13) > : H 4 : I 41 : P 55 : snmp_counter:tcp_passive(.1.3..6.0): 0 -> > buffer(): 10 (time P:71.5 | 0.19) > : H 4 : I 41 : P 60 : snmp_counter:load_average_5(.1.3..3.2): 0= .13 -> > buffer(): 11 (time P:2.81 | 0.19) > : H 4 : I 41 : P 70 : snmp_counter:load_average_15(.1.3..3.3): = 0.10 -> > buffer(): 12 (time P:0.97 | 0.18) > : H 4 : I 41 : P 75 : snmp_counter:mail_total(.1.3..1.1): "1179= " -> > buffer(): 13 (time P:173.07 | 0.15) > : H 4 : I 41 : P 80 : no_poller(): 0 -> rrd(*): cpu_user_ticks:= 43094 - > cpu_idle_ticks:6540643 - cpu_nice_ticks:1305 - cpu_system_ticks:79668 - > load_average_1:0.06 - load_average_5:0.13 - load_average_15:0.10 - num_= users:0 > - num_procs:261 - tcp_active:46088 - tcp_passive:0 - tcp_established:68= (time > P:0.19 | 117.95) > : H 4 : I 41 : P LPD : last_poll_date(): 1094153380 -> > db(last_poll_date): 1 (time P:0.21 | 2.29) > : H 4 : Poller End, Total Time: 392.6 msec. > > The problem is at the no_poller() line, no value for mail_total, but I = don't see > what's wrong in my new poller definition. > > Thanks for your help. > Pierre > > |
From: Javier S. <ja...@jf...> - 2004-09-02 22:09:50
|
Pierre, You should look into the shell script. Its returning the type as STRING instead of INTEGER. To convert it you have to do $value = str_replace("\"","",$value); Javier Pierre-Damien Gatouillat wrote: > Thanks for your reply. But my script returns, for example 872, but the snmp > function return "872" : > [root@relay2 engine]# snmpwalk -v 1 192.168.4.20 public .1.3.6.1.4.1.2021.51 > enterprises.ucdavis.51.1.1 = 1 > enterprises.ucdavis.51.2.1 = "mail" > enterprises.ucdavis.51.3.1 = "/usr/local/bin/stat_mail.sh" > enterprises.ucdavis.51.100.1 = 0 > enterprises.ucdavis.51.101.1 = "872" > enterprises.ucdavis.51.101.2 = "21" > enterprises.ucdavis.51.101.3 = "19" > enterprises.ucdavis.51.102.1 = 0 > > I try this : > if (is_string($value)) settype ($value, "integer"); > > But it doesn't work. > > Do you know how we can remove the quotes ? > > Pierre > > _________________ > Hi Pierre, > > Everything is fine. > > Except that the return value is a string, and it has to be a number. > > You don't see it in the RRD line because its not a number. > > So, you have 3 options. > > - Modify the stat_mail.sh to return an INTEGER without the " " > - Create a poller to get this OID and convert it to integer. > - Modify the snmp_counter poller to convert the return value to integer > (remove the quotes) > > I think the first option is the best. > > I hope this helps. > > Javier > > Pierre-Damien Gatouillat wrote: > > >>Hello everybody, >> >>I have had a new poller item like that : >>ID Description Name (Match RRD Struct DS) Poller Command > > (file) > >>Parameters >>10001 Mail Accounting Total mail_delete snmp_counter >> .1.3.6.1.4.1.2021.51.101.1 >> >>The OID is a private one and exec un shell script via snmpd.conf : >>exec .1.3.6.1.4.1.2021.51 mail /usr/local/bin/stat_mail.sh >> >>This OID works good and return a value : #snmpwalk -v 1 localhost public >>.1.3.6.1.4.1.2021.51.101.1 : >>enterprises.ucdavis.51.101.1 = "1171" >> >>Next add this poller to "Linux/Unix System Info" Interface Types : >>ID Description Internal Name Position Interface > > Type Field > >>Type Show Overwritable Tracked Default Value >> 10007 Mails Total mail_total 130 Linux/Unix System > > Info RRDTool > >>DS Never Type: Gauge Min: 0 Max: 10000 or Use for > > Max > >>Next add an entry in "Poller/Backend Grouping" in Linux/Unix Host: >> 10006 Linux/Unix Host 75 Mail Accounting Total > > Temporal Buffer > >>And a graphe type : >> 10001 Mails Linux/Unix System Info mail_stats > > 500 175 > >>0 0 >> >>And a new php script : mail_stats.inc.php with a copy of the rtt graph, change >>the value rtt by mail_stats. >> >>When I run php -q poller.php 4 41, the rrd file is created, I can see a graph, >>but after new polling, the rrd is not change, and no value in it (rrdtool dump >>interface-41-12.rrd) >>THe result of the poller : >>bash-2.05$ php -q poller.php 4 41 >> : H 4 : Poller Start : 14 Items. >> : H 4 : I 41 : P 10 : snmp_counter:cpu_nice_ticks(.1.3..1.0): 1305 -> >>buffer(): 1 (time P:3.19 | 0.34) >> : H 4 : I 41 : P 15 : snmp_counter:num_procs(.1.3..6.0): 261 -> >>buffer(): 2 (time P:1.4 | 0.12) >> : H 4 : I 41 : P 20 : snmp_counter:cpu_user_ticks(.1.3..0.0): 43094 -> >>buffer(): 3 (time P:1.12 | 0.12) >> : H 4 : I 41 : P 25 : snmp_counter:num_users(.1.3..5.0): 0 -> > > buffer(): > >>4 (time P:0.92 | 0.11) >> : H 4 : I 41 : P 30 : snmp_counter:cpu_idle_ticks(.1.3..3.0): 6540643 > > -> > >>buffer(): 5 (time P:1.08 | 0.11) >> : H 4 : I 41 : P 35 : snmp_counter:tcp_established(.1.3..9.0): 68 -> >>buffer(): 6 (time P:0.84 | 0.11) >> : H 4 : I 41 : P 40 : snmp_counter:cpu_system_ticks(.1.3..2.0): 79668 > > -> > >>buffer(): 7 (time P:1.08 | 0.11) >> : H 4 : I 41 : P 45 : snmp_counter:tcp_active(.1.3..5.0): 46088 -> >>buffer(): 8 (time P:0.71 | 0.11) >> : H 4 : I 41 : P 50 : snmp_counter:load_average_1(.1.3..3.1): 0.06 -> >>buffer(): 9 (time P:0.84 | 0.13) >> : H 4 : I 41 : P 55 : snmp_counter:tcp_passive(.1.3..6.0): 0 -> >>buffer(): 10 (time P:71.5 | 0.19) >> : H 4 : I 41 : P 60 : snmp_counter:load_average_5(.1.3..3.2): 0.13 -> >>buffer(): 11 (time P:2.81 | 0.19) >> : H 4 : I 41 : P 70 : snmp_counter:load_average_15(.1.3..3.3): 0.10 -> >>buffer(): 12 (time P:0.97 | 0.18) >> : H 4 : I 41 : P 75 : snmp_counter:mail_total(.1.3..1.1): "1179" -> >>buffer(): 13 (time P:173.07 | 0.15) >> : H 4 : I 41 : P 80 : no_poller(): 0 -> rrd(*): cpu_user_ticks:43094 - >>cpu_idle_ticks:6540643 - cpu_nice_ticks:1305 - cpu_system_ticks:79668 - >>load_average_1:0.06 - load_average_5:0.13 - load_average_15:0.10 - num_users:0 >>- num_procs:261 - tcp_active:46088 - tcp_passive:0 - tcp_established:68 (time >>P:0.19 | 117.95) >> : H 4 : I 41 : P LPD : last_poll_date(): 1094153380 -> >>db(last_poll_date): 1 (time P:0.21 | 2.29) >> : H 4 : Poller End, Total Time: 392.6 msec. >> >>The problem is at the no_poller() line, no value for mail_total, but I don't > > see > >>what's wrong in my new poller definition. >> >>Thanks for your help. >>Pierre >> >> > > > > ------------------------------------------------------- > This SF.Net email is sponsored by BEA Weblogic Workshop > FREE Java Enterprise J2EE developer tools! > Get your free copy of BEA WebLogic Workshop 8.1 today. > http://ads.osdn.com/?ad_idP47&alloc_id808&op=click > _______________________________________________ > jffnms-users mailing list > jff...@li... > https://lists.sourceforge.net/lists/listinfo/jffnms-users -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: Pierre-Damien G. <pd...@fr...> - 2004-09-07 21:11:36
|
Hi, My shell script is : #!/bin/bash DATE=`date +%Y%m%d` LOG_TEMP="/tmp/log.imss" grep "Final action" /usr/local/trend/imss/log/log.imss.$DATE* > $LOG_TEMP TOTAL=`wc -l $LOG_TEMP | awk {'print $1'}` DELETE=`grep Delete $LOG_TEMP | wc -l` QUARANTINE=`grep Quarantine $LOG_TEMP | wc -l` echo $TOTAL echo $DELETE echo $QUARANTINE I don't know how I can return an integer value... For the momemt, I put $value = str_replace("\"","",$value) in the snmp_counter poller, and it works. (I'm very happy :o) If you or someone else know how have an integer with this script, it woulb be nice. I try also with a perl script like this : #!/usr/bin/perl $uptime = `uptime`; $uptime =~ /up (.*?) day/; $up = int($1); print "$up\n"; But, same problem with snmpwalk, and the variable $up is an integer : cerebellum:~# snmpwalk -v 1 localhost public .1.3.6.1.4.1.2021.51 enterprises.ucdavis.51.1.1 = 1 enterprises.ucdavis.51.2.1 = "uptime" enterprises.ucdavis.51.3.1 = "/usr/local/mrtg/scripts/uptime.pl" enterprises.ucdavis.51.100.1 = 0 enterprises.ucdavis.51.101.1 = "53" The quotes are always here... I think there is a problem with the exec function of ucd-snmp. I have not try with net-snmp. I will try to write a mini-howto to have graph like uptime, apache stats (hits, trafic), dns requests, ... with this method. Pierre Javier Szyszlican a écrit : > Pierre, > > You should look into the shell script. > Its returning the type as STRING instead of INTEGER. > > To convert it you have to do > $value = str_replace("\"","",$value); > > Javier > > Pierre-Damien Gatouillat wrote: > >> Thanks for your reply. But my script returns, for example 872, but the >> snmp >> function return "872" : >> [root@relay2 engine]# snmpwalk -v 1 192.168.4.20 public >> .1.3.6.1.4.1.2021.51 >> enterprises.ucdavis.51.1.1 = 1 >> enterprises.ucdavis.51.2.1 = "mail" >> enterprises.ucdavis.51.3.1 = "/usr/local/bin/stat_mail.sh" >> enterprises.ucdavis.51.100.1 = 0 >> enterprises.ucdavis.51.101.1 = "872" >> enterprises.ucdavis.51.101.2 = "21" >> enterprises.ucdavis.51.101.3 = "19" >> enterprises.ucdavis.51.102.1 = 0 >> >> I try this : >> if (is_string($value)) settype ($value, "integer"); >> >> But it doesn't work. >> >> Do you know how we can remove the quotes ? >> >> Pierre >> >> _________________ >> Hi Pierre, >> >> Everything is fine. >> >> Except that the return value is a string, and it has to be a number. >> >> You don't see it in the RRD line because its not a number. >> >> So, you have 3 options. >> >> - Modify the stat_mail.sh to return an INTEGER without the " " >> - Create a poller to get this OID and convert it to integer. >> - Modify the snmp_counter poller to convert the return value to >> integer >> (remove the quotes) >> >> I think the first option is the best. >> >> I hope this helps. >> >> Javier >> >> Pierre-Damien Gatouillat wrote: >> >> >>> Hello everybody, >>> >>> I have had a new poller item like that : >>> ID Description Name (Match RRD Struct DS) Poller >>> Command >> >> >> (file) >> >>> Parameters >>> 10001 Mail Accounting Total mail_delete snmp_counter >>> .1.3.6.1.4.1.2021.51.101.1 >>> >>> The OID is a private one and exec un shell script via snmpd.conf : >>> exec .1.3.6.1.4.1.2021.51 mail /usr/local/bin/stat_mail.sh >>> >>> This OID works good and return a value : #snmpwalk -v 1 localhost public >>> .1.3.6.1.4.1.2021.51.101.1 : >>> enterprises.ucdavis.51.101.1 = "1171" >>> >>> Next add this poller to "Linux/Unix System Info" Interface Types : >>> ID Description Internal Name Position >>> Interface >> >> >> Type Field >> >>> Type Show Overwritable Tracked Default Value >>> 10007 Mails Total mail_total 130 Linux/Unix >>> System >> >> >> Info RRDTool >> >>> DS Never Type: Gauge Min: 0 Max: 10000 or >>> Use for >> >> >> Max >> >>> Next add an entry in "Poller/Backend Grouping" in Linux/Unix Host: >>> 10006 Linux/Unix Host 75 Mail Accounting Total >> >> >> Temporal Buffer >> >>> And a graphe type : >>> 10001 Mails Linux/Unix System Info >>> mail_stats >> >> >> 500 175 >> >>> 0 0 >>> >>> And a new php script : mail_stats.inc.php with a copy of the rtt >>> graph, change >>> the value rtt by mail_stats. >>> >>> When I run php -q poller.php 4 41, the rrd file is created, I can see >>> a graph, >>> but after new polling, the rrd is not change, and no value in it >>> (rrdtool dump >>> interface-41-12.rrd) >>> THe result of the poller : >>> bash-2.05$ php -q poller.php 4 41 >>> : H 4 : Poller Start : 14 Items. >>> : H 4 : I 41 : P 10 : snmp_counter:cpu_nice_ticks(.1.3..1.0): >>> 1305 -> >>> buffer(): 1 (time P:3.19 | 0.34) >>> : H 4 : I 41 : P 15 : snmp_counter:num_procs(.1.3..6.0): 261 -> >>> buffer(): 2 (time P:1.4 | 0.12) >>> : H 4 : I 41 : P 20 : snmp_counter:cpu_user_ticks(.1.3..0.0): >>> 43094 -> >>> buffer(): 3 (time P:1.12 | 0.12) >>> : H 4 : I 41 : P 25 : snmp_counter:num_users(.1.3..5.0): 0 -> >> >> >> buffer(): >> >>> 4 (time P:0.92 | 0.11) >>> : H 4 : I 41 : P 30 : snmp_counter:cpu_idle_ticks(.1.3..3.0): >>> 6540643 >> >> >> -> >> >>> buffer(): 5 (time P:1.08 | 0.11) >>> : H 4 : I 41 : P 35 : snmp_counter:tcp_established(.1.3..9.0): >>> 68 -> >>> buffer(): 6 (time P:0.84 | 0.11) >>> : H 4 : I 41 : P 40 : >>> snmp_counter:cpu_system_ticks(.1.3..2.0): 79668 >> >> >> -> >> >>> buffer(): 7 (time P:1.08 | 0.11) >>> : H 4 : I 41 : P 45 : snmp_counter:tcp_active(.1.3..5.0): >>> 46088 -> >>> buffer(): 8 (time P:0.71 | 0.11) >>> : H 4 : I 41 : P 50 : snmp_counter:load_average_1(.1.3..3.1): >>> 0.06 -> >>> buffer(): 9 (time P:0.84 | 0.13) >>> : H 4 : I 41 : P 55 : snmp_counter:tcp_passive(.1.3..6.0): 0 -> >>> buffer(): 10 (time P:71.5 | 0.19) >>> : H 4 : I 41 : P 60 : snmp_counter:load_average_5(.1.3..3.2): >>> 0.13 -> >>> buffer(): 11 (time P:2.81 | 0.19) >>> : H 4 : I 41 : P 70 : snmp_counter:load_average_15(.1.3..3.3): >>> 0.10 -> >>> buffer(): 12 (time P:0.97 | 0.18) >>> : H 4 : I 41 : P 75 : snmp_counter:mail_total(.1.3..1.1): >>> "1179" -> >>> buffer(): 13 (time P:173.07 | 0.15) >>> : H 4 : I 41 : P 80 : no_poller(): 0 -> rrd(*): >>> cpu_user_ticks:43094 - >>> cpu_idle_ticks:6540643 - cpu_nice_ticks:1305 - cpu_system_ticks:79668 - >>> load_average_1:0.06 - load_average_5:0.13 - load_average_15:0.10 - >>> num_users:0 >>> - num_procs:261 - tcp_active:46088 - tcp_passive:0 - >>> tcp_established:68 (time >>> P:0.19 | 117.95) >>> : H 4 : I 41 : P LPD : last_poll_date(): 1094153380 -> >>> db(last_poll_date): 1 (time P:0.21 | 2.29) >>> : H 4 : Poller End, Total Time: 392.6 msec. >>> >>> The problem is at the no_poller() line, no value for mail_total, but >>> I don't >> >> >> see >> >>> what's wrong in my new poller definition. >>> >>> Thanks for your help. >>> Pierre >>> >>> >> >> >> >> ------------------------------------------------------- >> This SF.Net email is sponsored by BEA Weblogic Workshop >> FREE Java Enterprise J2EE developer tools! >> Get your free copy of BEA WebLogic Workshop 8.1 today. >> http://ads.osdn.com/?ad_idP47&alloc_id808&op=click >> _______________________________________________ >> jffnms-users mailing list >> jff...@li... >> https://lists.sourceforge.net/lists/listinfo/jffnms-users > > |
From: Javier S. <ja...@jf...> - 2004-09-07 21:23:55
|
Hi Pierre, Yes, this seems to be a bug/feature of ucd-snmp maybe net-snmp too. I use "pass" instead of exec that allows me to specify the type, but its = more=20 complex. And for the Apache Hits, we made an interface type for it, its on todays=20 nightly, and it will be included in 0.7.9. This uses the /server-status? = URL to=20 gather information from the remote apache host. Can you try it? to see if it does the same as you want. Javier Pierre-Damien Gatouillat wrote: > Hi, >=20 > My shell script is : > #!/bin/bash >=20 > DATE=3D`date +%Y%m%d` > LOG_TEMP=3D"/tmp/log.imss" > grep "Final action" /usr/local/trend/imss/log/log.imss.$DATE* > $LOG_TE= MP > TOTAL=3D`wc -l $LOG_TEMP | awk {'print $1'}` > DELETE=3D`grep Delete $LOG_TEMP | wc -l` > QUARANTINE=3D`grep Quarantine $LOG_TEMP | wc -l` >=20 > echo $TOTAL > echo $DELETE > echo $QUARANTINE >=20 > I don't know how I can return an integer value... >=20 > For the momemt, I put $value =3D str_replace("\"","",$value) in the=20 > snmp_counter poller, and it works. (I'm very happy :o) >=20 > If you or someone else know how have an integer with this script, it=20 > woulb be nice. > I try also with a perl script like this : > #!/usr/bin/perl >=20 > $uptime =3D `uptime`; >=20 > $uptime =3D~ /up (.*?) day/; > $up =3D int($1); >=20 > print "$up\n"; >=20 > But, same problem with snmpwalk, and the variable $up is an integer : > cerebellum:~# snmpwalk -v 1 localhost public .1.3.6.1.4.1.2021.51 > enterprises.ucdavis.51.1.1 =3D 1 > enterprises.ucdavis.51.2.1 =3D "uptime" > enterprises.ucdavis.51.3.1 =3D "/usr/local/mrtg/scripts/uptime.pl" > enterprises.ucdavis.51.100.1 =3D 0 > enterprises.ucdavis.51.101.1 =3D "53" >=20 > The quotes are always here... I think there is a problem with the exec=20 > function of ucd-snmp. I have not try with net-snmp. >=20 > I will try to write a mini-howto to have graph like uptime, apache stat= s=20 > (hits, trafic), dns requests, ... with this method. >=20 > Pierre >=20 >=20 >=20 > Javier Szyszlican a =E9crit : >=20 >> Pierre, >> >> You should look into the shell script. >> Its returning the type as STRING instead of INTEGER. >> >> To convert it you have to do >> $value =3D str_replace("\"","",$value); >> >> Javier >> >> Pierre-Damien Gatouillat wrote: >> >>> Thanks for your reply. But my script returns, for example 872, but=20 >>> the snmp >>> function return "872" : >>> [root@relay2 engine]# snmpwalk -v 1 192.168.4.20 public =20 >>> .1.3.6.1.4.1.2021.51 >>> enterprises.ucdavis.51.1.1 =3D 1 >>> enterprises.ucdavis.51.2.1 =3D "mail" >>> enterprises.ucdavis.51.3.1 =3D "/usr/local/bin/stat_mail.sh" >>> enterprises.ucdavis.51.100.1 =3D 0 >>> enterprises.ucdavis.51.101.1 =3D "872" >>> enterprises.ucdavis.51.101.2 =3D "21" >>> enterprises.ucdavis.51.101.3 =3D "19" >>> enterprises.ucdavis.51.102.1 =3D 0 >>> >>> I try this : >>> if (is_string($value)) settype ($value, "integer"); >>> >>> But it doesn't work. >>> >>> Do you know how we can remove the quotes ? >>> >>> Pierre >>> >>> _________________ >>> Hi Pierre, >>> >>> Everything is fine. >>> >>> Except that the return value is a string, and it has to be a number. >>> >>> You don't see it in the RRD line because its not a number. >>> >>> So, you have 3 options. >>> >>> - Modify the stat_mail.sh to return an INTEGER without the " " >>> - Create a poller to get this OID and convert it to integer. >>> - Modify the snmp_counter poller to convert the return value to=20 >>> integer >>> (remove the quotes) >>> >>> I think the first option is the best. >>> >>> I hope this helps. >>> >>> Javier >>> >>> Pierre-Damien Gatouillat wrote: >>> >>> >>>> Hello everybody, >>>> >>>> I have had a new poller item like that : >>>> ID Description Name (Match RRD Struct DS) =20 >>>> Poller Command >>> >>> >>> >>> (file) >>> >>>> Parameters >>>> 10001 Mail Accounting Total mail_delete snmp_counter >>>> .1.3.6.1.4.1.2021.51.101.1 >>>> >>>> The OID is a private one and exec un shell script via snmpd.conf : >>>> exec .1.3.6.1.4.1.2021.51 mail /usr/local/bin/stat_mail.sh >>>> >>>> This OID works good and return a value : #snmpwalk -v 1 localhost=20 >>>> public >>>> .1.3.6.1.4.1.2021.51.101.1 : >>>> enterprises.ucdavis.51.101.1 =3D "1171" >>>> >>>> Next add this poller to "Linux/Unix System Info" Interface Types : >>>> ID Description Internal Name Position =20 >>>> Interface >>> >>> >>> >>> Type Field >>> >>>> Type Show Overwritable Tracked Default=20 >>>> Value >>>> 10007 Mails Total mail_total 130 Linux/Unix=20 >>>> System >>> >>> >>> >>> Info RRDTool >>> >>>> DS Never Type: Gauge Min: 0 Max: 10000 or=20 >>>> Use for >>> >>> >>> >>> Max >>> >>>> Next add an entry in "Poller/Backend Grouping" in Linux/Unix Host: >>>> 10006 Linux/Unix Host 75 Mail Accounting Total = =20 >>> >>> >>> >>> Temporal Buffer >>> >>>> And a graphe type : >>>> 10001 Mails Linux/Unix System Info =20 >>>> mail_stats =20 >>> >>> >>> >>> 500 175 >>> >>>> 0 0 >>>> >>>> And a new php script : mail_stats.inc.php with a copy of the rtt=20 >>>> graph, change >>>> the value rtt by mail_stats. >>>> >>>> When I run php -q poller.php 4 41, the rrd file is created, I can=20 >>>> see a graph, >>>> but after new polling, the rrd is not change, and no value in it=20 >>>> (rrdtool dump >>>> interface-41-12.rrd) >>>> THe result of the poller : >>>> bash-2.05$ php -q poller.php 4 41 >>>> : H 4 : Poller Start : 14 Items. >>>> : H 4 : I 41 : P 10 : snmp_counter:cpu_nice_ticks(.1.3..1.0):= =20 >>>> 1305 -> >>>> buffer(): 1 (time P:3.19 | 0.34) >>>> : H 4 : I 41 : P 15 : snmp_counter:num_procs(.1.3..6.0): 261 = -> >>>> buffer(): 2 (time P:1.4 | 0.12) >>>> : H 4 : I 41 : P 20 : snmp_counter:cpu_user_ticks(.1.3..0.0):= =20 >>>> 43094 -> >>>> buffer(): 3 (time P:1.12 | 0.12) >>>> : H 4 : I 41 : P 25 : snmp_counter:num_users(.1.3..5.0): 0 -> >>> >>> >>> >>> buffer(): >>> >>>> 4 (time P:0.92 | 0.11) >>>> : H 4 : I 41 : P 30 : snmp_counter:cpu_idle_ticks(.1.3..3.0):= =20 >>>> 6540643 >>> >>> >>> >>> -> >>> >>>> buffer(): 5 (time P:1.08 | 0.11) >>>> : H 4 : I 41 : P 35 :=20 >>>> snmp_counter:tcp_established(.1.3..9.0): 68 -> >>>> buffer(): 6 (time P:0.84 | 0.11) >>>> : H 4 : I 41 : P 40 :=20 >>>> snmp_counter:cpu_system_ticks(.1.3..2.0): 79668 >>> >>> >>> >>> -> >>> >>>> buffer(): 7 (time P:1.08 | 0.11) >>>> : H 4 : I 41 : P 45 : snmp_counter:tcp_active(.1.3..5.0):=20 >>>> 46088 -> >>>> buffer(): 8 (time P:0.71 | 0.11) >>>> : H 4 : I 41 : P 50 : snmp_counter:load_average_1(.1.3..3.1):= =20 >>>> 0.06 -> >>>> buffer(): 9 (time P:0.84 | 0.13) >>>> : H 4 : I 41 : P 55 : snmp_counter:tcp_passive(.1.3..6.0): 0 = -> >>>> buffer(): 10 (time P:71.5 | 0.19) >>>> : H 4 : I 41 : P 60 : snmp_counter:load_average_5(.1.3..3.2):= =20 >>>> 0.13 -> >>>> buffer(): 11 (time P:2.81 | 0.19) >>>> : H 4 : I 41 : P 70 :=20 >>>> snmp_counter:load_average_15(.1.3..3.3): 0.10 -> >>>> buffer(): 12 (time P:0.97 | 0.18) >>>> : H 4 : I 41 : P 75 : snmp_counter:mail_total(.1.3..1.1):=20 >>>> "1179" -> >>>> buffer(): 13 (time P:173.07 | 0.15) >>>> : H 4 : I 41 : P 80 : no_poller(): 0 -> rrd(*):=20 >>>> cpu_user_ticks:43094 - >>>> cpu_idle_ticks:6540643 - cpu_nice_ticks:1305 - cpu_system_ticks:7966= 8 - >>>> load_average_1:0.06 - load_average_5:0.13 - load_average_15:0.10 -=20 >>>> num_users:0 >>>> - num_procs:261 - tcp_active:46088 - tcp_passive:0 -=20 >>>> tcp_established:68 (time >>>> P:0.19 | 117.95) >>>> : H 4 : I 41 : P LPD : last_poll_date(): 1094153380 -> >>>> db(last_poll_date): 1 (time P:0.21 | 2.29) >>>> : H 4 : Poller End, Total Time: 392.6 msec. >>>> >>>> The problem is at the no_poller() line, no value for mail_total, but= =20 >>>> I don't >>> >>> >>> >>> see >>> >>>> what's wrong in my new poller definition. >>>> >>>> Thanks for your help. >>>> Pierre >>>> >>>> >>> >>> >>> >>> ------------------------------------------------------- >>> This SF.Net email is sponsored by BEA Weblogic Workshop >>> FREE Java Enterprise J2EE developer tools! >>> Get your free copy of BEA WebLogic Workshop 8.1 today. >>> http://ads.osdn.com/?ad_idP47&alloc_id=10808&op=3Dclick >>> _______________________________________________ >>> jffnms-users mailing list >>> jff...@li... >>> https://lists.sourceforge.net/lists/listinfo/jffnms-users >> >> >> >=20 >=20 > ------------------------------------------------------- > This SF.Net email is sponsored by BEA Weblogic Workshop > FREE Java Enterprise J2EE developer tools! > Get your free copy of BEA WebLogic Workshop 8.1 today. > http://ads.osdn.com/?ad_id=3D5047&alloc_id=3D10808&op=3Dclick > _______________________________________________ > jffnms-users mailing list > jff...@li... > https://lists.sourceforge.net/lists/listinfo/jffnms-users --=20 =3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D= -=3D-=3D-=3D-=3D-=3D Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: Pierre-Damien G. <pd...@fr...> - 2004-09-07 21:40:13
|
I'm sorry, but I have no dev installation of jff to try it, maybe I try it on this week-end. The script I use is also based on the /server-status? URL (used with MRTG) #!/usr/bin/perl # depends on availability of status and extended status info from your # Apache webserver -- your httpd.conf needs to include something like the # following: (uncommented) #<Location /server-status> # SetHandler server-status # Order allow,deny # Allow from localhost #</Location> #ExtendedStatus On # can return hits or bytes (counters) @res = `lynx -dump http://localhost:80/server-status`; foreach $res (@res) { if ($res =~ /Server uptime: (.*)$/) { $up = $1; last } else { next } if ($res =~ /Server at/) { $server = $res; last } else { next } } @res = `lynx -dump http://localhost:80/server-status?auto`; foreach $res (@res) { if ($res =~ /Total Accesses: (\d+)/) { $d1 = $1; next } if ($res =~ /Total kBytes: (\d+)/) { $d2 = $1 * 1024; next } } $d1 = int($d1); $d2 = int($d2); if ($ARGV[0] eq "hits") { print "$d1\n"; print "$d1\n"; } elsif ($ARGV[0] eq "bytes") { print "$d2\n"; print "$d2\n"; } print "$up\n"; print "$server"; Javier Szyszlican a écrit : > Hi Pierre, > > Yes, this seems to be a bug/feature of ucd-snmp maybe net-snmp too. > > I use "pass" instead of exec that allows me to specify the type, but its > more complex. > > And for the Apache Hits, we made an interface type for it, its on todays > nightly, and it will be included in 0.7.9. This uses the /server-status? > URL to gather information from the remote apache host. > > Can you try it? to see if it does the same as you want. > > Javier > > Pierre-Damien Gatouillat wrote: > >> Hi, >> >> My shell script is : >> #!/bin/bash >> >> DATE=`date +%Y%m%d` >> LOG_TEMP="/tmp/log.imss" >> grep "Final action" /usr/local/trend/imss/log/log.imss.$DATE* > $LOG_TEMP >> TOTAL=`wc -l $LOG_TEMP | awk {'print $1'}` >> DELETE=`grep Delete $LOG_TEMP | wc -l` >> QUARANTINE=`grep Quarantine $LOG_TEMP | wc -l` >> >> echo $TOTAL >> echo $DELETE >> echo $QUARANTINE >> >> I don't know how I can return an integer value... >> >> For the momemt, I put $value = str_replace("\"","",$value) in the >> snmp_counter poller, and it works. (I'm very happy :o) >> >> If you or someone else know how have an integer with this script, it >> woulb be nice. >> I try also with a perl script like this : >> #!/usr/bin/perl >> >> $uptime = `uptime`; >> >> $uptime =~ /up (.*?) day/; >> $up = int($1); >> >> print "$up\n"; >> >> But, same problem with snmpwalk, and the variable $up is an integer : >> cerebellum:~# snmpwalk -v 1 localhost public .1.3.6.1.4.1.2021.51 >> enterprises.ucdavis.51.1.1 = 1 >> enterprises.ucdavis.51.2.1 = "uptime" >> enterprises.ucdavis.51.3.1 = "/usr/local/mrtg/scripts/uptime.pl" >> enterprises.ucdavis.51.100.1 = 0 >> enterprises.ucdavis.51.101.1 = "53" >> >> The quotes are always here... I think there is a problem with the exec >> function of ucd-snmp. I have not try with net-snmp. >> >> I will try to write a mini-howto to have graph like uptime, apache >> stats (hits, trafic), dns requests, ... with this method. >> >> Pierre >> >> >> >> Javier Szyszlican a écrit : >> >>> Pierre, >>> >>> You should look into the shell script. >>> Its returning the type as STRING instead of INTEGER. >>> >>> To convert it you have to do >>> $value = str_replace("\"","",$value); >>> >>> Javier >>> >>> Pierre-Damien Gatouillat wrote: >>> >>>> Thanks for your reply. But my script returns, for example 872, but >>>> the snmp >>>> function return "872" : >>>> [root@relay2 engine]# snmpwalk -v 1 192.168.4.20 public >>>> .1.3.6.1.4.1.2021.51 >>>> enterprises.ucdavis.51.1.1 = 1 >>>> enterprises.ucdavis.51.2.1 = "mail" >>>> enterprises.ucdavis.51.3.1 = "/usr/local/bin/stat_mail.sh" >>>> enterprises.ucdavis.51.100.1 = 0 >>>> enterprises.ucdavis.51.101.1 = "872" >>>> enterprises.ucdavis.51.101.2 = "21" >>>> enterprises.ucdavis.51.101.3 = "19" >>>> enterprises.ucdavis.51.102.1 = 0 >>>> >>>> I try this : >>>> if (is_string($value)) settype ($value, "integer"); >>>> >>>> But it doesn't work. >>>> >>>> Do you know how we can remove the quotes ? >>>> >>>> Pierre >>>> >>>> _________________ >>>> Hi Pierre, >>>> >>>> Everything is fine. >>>> >>>> Except that the return value is a string, and it has to be a number. >>>> >>>> You don't see it in the RRD line because its not a number. >>>> >>>> So, you have 3 options. >>>> >>>> - Modify the stat_mail.sh to return an INTEGER without the " " >>>> - Create a poller to get this OID and convert it to integer. >>>> - Modify the snmp_counter poller to convert the return value to >>>> integer >>>> (remove the quotes) >>>> >>>> I think the first option is the best. >>>> >>>> I hope this helps. >>>> >>>> Javier >>>> >>>> Pierre-Damien Gatouillat wrote: >>>> >>>> >>>>> Hello everybody, >>>>> >>>>> I have had a new poller item like that : >>>>> ID Description Name (Match RRD Struct DS) >>>>> Poller Command >>>> >>>> >>>> >>>> >>>> (file) >>>> >>>>> Parameters >>>>> 10001 Mail Accounting Total mail_delete snmp_counter >>>>> .1.3.6.1.4.1.2021.51.101.1 >>>>> >>>>> The OID is a private one and exec un shell script via snmpd.conf : >>>>> exec .1.3.6.1.4.1.2021.51 mail /usr/local/bin/stat_mail.sh >>>>> >>>>> This OID works good and return a value : #snmpwalk -v 1 localhost >>>>> public >>>>> .1.3.6.1.4.1.2021.51.101.1 : >>>>> enterprises.ucdavis.51.101.1 = "1171" >>>>> >>>>> Next add this poller to "Linux/Unix System Info" Interface Types : >>>>> ID Description Internal Name Position >>>>> Interface >>>> >>>> >>>> >>>> >>>> Type Field >>>> >>>>> Type Show Overwritable Tracked Default >>>>> Value >>>>> 10007 Mails Total mail_total 130 Linux/Unix >>>>> System >>>> >>>> >>>> >>>> >>>> Info RRDTool >>>> >>>>> DS Never Type: Gauge Min: 0 Max: 10000 or >>>>> Use for >>>> >>>> >>>> >>>> >>>> Max >>>> >>>>> Next add an entry in "Poller/Backend Grouping" in Linux/Unix Host: >>>>> 10006 Linux/Unix Host 75 Mail Accounting Total >>>> >>>> >>>> >>>> >>>> Temporal Buffer >>>> >>>>> And a graphe type : >>>>> 10001 Mails Linux/Unix System Info >>>>> mail_stats >>>> >>>> >>>> >>>> >>>> 500 175 >>>> >>>>> 0 0 >>>>> >>>>> And a new php script : mail_stats.inc.php with a copy of the rtt >>>>> graph, change >>>>> the value rtt by mail_stats. >>>>> >>>>> When I run php -q poller.php 4 41, the rrd file is created, I can >>>>> see a graph, >>>>> but after new polling, the rrd is not change, and no value in it >>>>> (rrdtool dump >>>>> interface-41-12.rrd) >>>>> THe result of the poller : >>>>> bash-2.05$ php -q poller.php 4 41 >>>>> : H 4 : Poller Start : 14 Items. >>>>> : H 4 : I 41 : P 10 : >>>>> snmp_counter:cpu_nice_ticks(.1.3..1.0): 1305 -> >>>>> buffer(): 1 (time P:3.19 | 0.34) >>>>> : H 4 : I 41 : P 15 : snmp_counter:num_procs(.1.3..6.0): 261 -> >>>>> buffer(): 2 (time P:1.4 | 0.12) >>>>> : H 4 : I 41 : P 20 : >>>>> snmp_counter:cpu_user_ticks(.1.3..0.0): 43094 -> >>>>> buffer(): 3 (time P:1.12 | 0.12) >>>>> : H 4 : I 41 : P 25 : snmp_counter:num_users(.1.3..5.0): 0 -> >>>> >>>> >>>> >>>> >>>> buffer(): >>>> >>>>> 4 (time P:0.92 | 0.11) >>>>> : H 4 : I 41 : P 30 : >>>>> snmp_counter:cpu_idle_ticks(.1.3..3.0): 6540643 >>>> >>>> >>>> >>>> >>>> -> >>>> >>>>> buffer(): 5 (time P:1.08 | 0.11) >>>>> : H 4 : I 41 : P 35 : >>>>> snmp_counter:tcp_established(.1.3..9.0): 68 -> >>>>> buffer(): 6 (time P:0.84 | 0.11) >>>>> : H 4 : I 41 : P 40 : >>>>> snmp_counter:cpu_system_ticks(.1.3..2.0): 79668 >>>> >>>> >>>> >>>> >>>> -> >>>> >>>>> buffer(): 7 (time P:1.08 | 0.11) >>>>> : H 4 : I 41 : P 45 : snmp_counter:tcp_active(.1.3..5.0): >>>>> 46088 -> >>>>> buffer(): 8 (time P:0.71 | 0.11) >>>>> : H 4 : I 41 : P 50 : >>>>> snmp_counter:load_average_1(.1.3..3.1): 0.06 -> >>>>> buffer(): 9 (time P:0.84 | 0.13) >>>>> : H 4 : I 41 : P 55 : snmp_counter:tcp_passive(.1.3..6.0): 0 -> >>>>> buffer(): 10 (time P:71.5 | 0.19) >>>>> : H 4 : I 41 : P 60 : >>>>> snmp_counter:load_average_5(.1.3..3.2): 0.13 -> >>>>> buffer(): 11 (time P:2.81 | 0.19) >>>>> : H 4 : I 41 : P 70 : >>>>> snmp_counter:load_average_15(.1.3..3.3): 0.10 -> >>>>> buffer(): 12 (time P:0.97 | 0.18) >>>>> : H 4 : I 41 : P 75 : snmp_counter:mail_total(.1.3..1.1): >>>>> "1179" -> >>>>> buffer(): 13 (time P:173.07 | 0.15) >>>>> : H 4 : I 41 : P 80 : no_poller(): 0 -> rrd(*): >>>>> cpu_user_ticks:43094 - >>>>> cpu_idle_ticks:6540643 - cpu_nice_ticks:1305 - >>>>> cpu_system_ticks:79668 - >>>>> load_average_1:0.06 - load_average_5:0.13 - load_average_15:0.10 - >>>>> num_users:0 >>>>> - num_procs:261 - tcp_active:46088 - tcp_passive:0 - >>>>> tcp_established:68 (time >>>>> P:0.19 | 117.95) >>>>> : H 4 : I 41 : P LPD : last_poll_date(): 1094153380 -> >>>>> db(last_poll_date): 1 (time P:0.21 | 2.29) >>>>> : H 4 : Poller End, Total Time: 392.6 msec. >>>>> >>>>> The problem is at the no_poller() line, no value for mail_total, >>>>> but I don't >>>> >>>> >>>> >>>> >>>> see >>>> >>>>> what's wrong in my new poller definition. >>>>> >>>>> Thanks for your help. >>>>> Pierre >>>>> >>>>> >>>> >>>> >>>> >>>> ------------------------------------------------------- >>>> This SF.Net email is sponsored by BEA Weblogic Workshop >>>> FREE Java Enterprise J2EE developer tools! >>>> Get your free copy of BEA WebLogic Workshop 8.1 today. >>>> http://ads.osdn.com/?ad_idP47&alloc_id808&op=click >>>> _______________________________________________ >>>> jffnms-users mailing list >>>> jff...@li... >>>> https://lists.sourceforge.net/lists/listinfo/jffnms-users >>> >>> >>> >>> >> >> >> ------------------------------------------------------- >> This SF.Net email is sponsored by BEA Weblogic Workshop >> FREE Java Enterprise J2EE developer tools! >> Get your free copy of BEA WebLogic Workshop 8.1 today. >> http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click >> _______________________________________________ >> jffnms-users mailing list >> jff...@li... >> https://lists.sourceforge.net/lists/listinfo/jffnms-users > > |
From: Javier S. <ja...@jf...> - 2004-09-07 21:44:00
|
Pierre, Well, we 'implemented' that in PHP inside JFFNMS. So JFFNMS does the HTTP GET to apache, parses that return info and feeds = the graphs. Javier Pierre-Damien Gatouillat wrote: > I'm sorry, but I have no dev installation of jff to try it, maybe I try= =20 > it on this week-end. >=20 > The script I use is also based on the /server-status? URL > (used with MRTG) > #!/usr/bin/perl >=20 > # depends on availability of status and extended status info from your > # Apache webserver -- your httpd.conf needs to include something like t= he > # following: (uncommented) > #<Location /server-status> > # SetHandler server-status > # Order allow,deny > # Allow from localhost > #</Location> > #ExtendedStatus On >=20 > # can return hits or bytes (counters) >=20 > @res =3D `lynx -dump http://localhost:80/server-status`; >=20 > foreach $res (@res) { > if ($res =3D~ /Server uptime: (.*)$/) { $up =3D $1; last } else { n= ext } > if ($res =3D~ /Server at/) { $server =3D $res; last } else { next } > } >=20 > @res =3D `lynx -dump http://localhost:80/server-status?auto`; >=20 > foreach $res (@res) { > if ($res =3D~ /Total Accesses: (\d+)/) { $d1 =3D $1; next } > if ($res =3D~ /Total kBytes: (\d+)/) { $d2 =3D $1 * 1024; next } > } >=20 > $d1 =3D int($d1); > $d2 =3D int($d2); >=20 > if ($ARGV[0] eq "hits") { > print "$d1\n"; > print "$d1\n"; > } elsif ($ARGV[0] eq "bytes") { > print "$d2\n"; > print "$d2\n"; > } >=20 > print "$up\n"; > print "$server"; >=20 > Javier Szyszlican a =E9crit : >=20 >> Hi Pierre, >> >> Yes, this seems to be a bug/feature of ucd-snmp maybe net-snmp too. >> >> I use "pass" instead of exec that allows me to specify the type, but=20 >> its more complex. >> >> And for the Apache Hits, we made an interface type for it, its on=20 >> todays nightly, and it will be included in 0.7.9. This uses the=20 >> /server-status? URL to gather information from the remote apache host. >> >> Can you try it? to see if it does the same as you want. >> >> Javier >> >> Pierre-Damien Gatouillat wrote: >> >>> Hi, >>> >>> My shell script is : >>> #!/bin/bash >>> >>> DATE=3D`date +%Y%m%d` >>> LOG_TEMP=3D"/tmp/log.imss" >>> grep "Final action" /usr/local/trend/imss/log/log.imss.$DATE* >=20 >>> $LOG_TEMP >>> TOTAL=3D`wc -l $LOG_TEMP | awk {'print $1'}` >>> DELETE=3D`grep Delete $LOG_TEMP | wc -l` >>> QUARANTINE=3D`grep Quarantine $LOG_TEMP | wc -l` >>> >>> echo $TOTAL >>> echo $DELETE >>> echo $QUARANTINE >>> >>> I don't know how I can return an integer value... >>> >>> For the momemt, I put $value =3D str_replace("\"","",$value) in the=20 >>> snmp_counter poller, and it works. (I'm very happy :o) >>> >>> If you or someone else know how have an integer with this script, it=20 >>> woulb be nice. >>> I try also with a perl script like this : >>> #!/usr/bin/perl >>> >>> $uptime =3D `uptime`; >>> >>> $uptime =3D~ /up (.*?) day/; >>> $up =3D int($1); >>> >>> print "$up\n"; >>> >>> But, same problem with snmpwalk, and the variable $up is an integer : >>> cerebellum:~# snmpwalk -v 1 localhost public .1.3.6.1.4.1.2021.51 >>> enterprises.ucdavis.51.1.1 =3D 1 >>> enterprises.ucdavis.51.2.1 =3D "uptime" >>> enterprises.ucdavis.51.3.1 =3D "/usr/local/mrtg/scripts/uptime.pl" >>> enterprises.ucdavis.51.100.1 =3D 0 >>> enterprises.ucdavis.51.101.1 =3D "53" >>> >>> The quotes are always here... I think there is a problem with the=20 >>> exec function of ucd-snmp. I have not try with net-snmp. >>> >>> I will try to write a mini-howto to have graph like uptime, apache=20 >>> stats (hits, trafic), dns requests, ... with this method. >>> >>> Pierre >>> >>> >>> >>> Javier Szyszlican a =E9crit : >>> >>>> Pierre, >>>> >>>> You should look into the shell script. >>>> Its returning the type as STRING instead of INTEGER. >>>> >>>> To convert it you have to do >>>> $value =3D str_replace("\"","",$value); >>>> >>>> Javier >>>> >>>> Pierre-Damien Gatouillat wrote: >>>> >>>>> Thanks for your reply. But my script returns, for example 872, but=20 >>>>> the snmp >>>>> function return "872" : >>>>> [root@relay2 engine]# snmpwalk -v 1 192.168.4.20 public =20 >>>>> .1.3.6.1.4.1.2021.51 >>>>> enterprises.ucdavis.51.1.1 =3D 1 >>>>> enterprises.ucdavis.51.2.1 =3D "mail" >>>>> enterprises.ucdavis.51.3.1 =3D "/usr/local/bin/stat_mail.sh" >>>>> enterprises.ucdavis.51.100.1 =3D 0 >>>>> enterprises.ucdavis.51.101.1 =3D "872" >>>>> enterprises.ucdavis.51.101.2 =3D "21" >>>>> enterprises.ucdavis.51.101.3 =3D "19" >>>>> enterprises.ucdavis.51.102.1 =3D 0 >>>>> >>>>> I try this : >>>>> if (is_string($value)) settype ($value, "integer"); >>>>> >>>>> But it doesn't work. >>>>> >>>>> Do you know how we can remove the quotes ? >>>>> >>>>> Pierre >>>>> >>>>> _________________ >>>>> Hi Pierre, >>>>> >>>>> Everything is fine. >>>>> >>>>> Except that the return value is a string, and it has to be a number= . >>>>> >>>>> You don't see it in the RRD line because its not a number. >>>>> >>>>> So, you have 3 options. >>>>> >>>>> - Modify the stat_mail.sh to return an INTEGER without the " " >>>>> - Create a poller to get this OID and convert it to integer. >>>>> - Modify the snmp_counter poller to convert the return value to= =20 >>>>> integer >>>>> (remove the quotes) >>>>> >>>>> I think the first option is the best. >>>>> >>>>> I hope this helps. >>>>> >>>>> Javier >>>>> >>>>> Pierre-Damien Gatouillat wrote: >>>>> >>>>> >>>>>> Hello everybody, >>>>>> >>>>>> I have had a new poller item like that : >>>>>> ID Description Name (Match RRD Struct DS) =20 >>>>>> Poller Command >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> (file) >>>>> >>>>>> Parameters >>>>>> 10001 Mail Accounting Total mail_delete snmp_counter >>>>>> .1.3.6.1.4.1.2021.51.101.1 >>>>>> >>>>>> The OID is a private one and exec un shell script via snmpd.conf : >>>>>> exec .1.3.6.1.4.1.2021.51 mail /usr/local/bin/stat_mail.sh >>>>>> >>>>>> This OID works good and return a value : #snmpwalk -v 1 localhost=20 >>>>>> public >>>>>> .1.3.6.1.4.1.2021.51.101.1 : >>>>>> enterprises.ucdavis.51.101.1 =3D "1171" >>>>>> >>>>>> Next add this poller to "Linux/Unix System Info" Interface Types : >>>>>> ID Description Internal Name Position =20 >>>>>> Interface >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Type Field >>>>> >>>>>> Type Show Overwritable Tracked Default= =20 >>>>>> Value >>>>>> 10007 Mails Total mail_total 130 =20 >>>>>> Linux/Unix System >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Info RRDTool >>>>> >>>>>> DS Never Type: Gauge Min: 0 Max: 10000=20 >>>>>> or Use for >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Max >>>>> >>>>>> Next add an entry in "Poller/Backend Grouping" in Linux/Unix Host: >>>>>> 10006 Linux/Unix Host 75 Mail Accounting Total = =20 >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Temporal Buffer >>>>> >>>>>> And a graphe type : >>>>>> 10001 Mails Linux/Unix System Info =20 >>>>>> mail_stats =20 >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> 500 175 >>>>> >>>>>> 0 0 >>>>>> >>>>>> And a new php script : mail_stats.inc.php with a copy of the rtt=20 >>>>>> graph, change >>>>>> the value rtt by mail_stats. >>>>>> >>>>>> When I run php -q poller.php 4 41, the rrd file is created, I can=20 >>>>>> see a graph, >>>>>> but after new polling, the rrd is not change, and no value in it=20 >>>>>> (rrdtool dump >>>>>> interface-41-12.rrd) >>>>>> THe result of the poller : >>>>>> bash-2.05$ php -q poller.php 4 41 >>>>>> : H 4 : Poller Start : 14 Items. >>>>>> : H 4 : I 41 : P 10 :=20 >>>>>> snmp_counter:cpu_nice_ticks(.1.3..1.0): 1305 -> >>>>>> buffer(): 1 (time P:3.19 | 0.34) >>>>>> : H 4 : I 41 : P 15 : snmp_counter:num_procs(.1.3..6.0):=20 >>>>>> 261 -> >>>>>> buffer(): 2 (time P:1.4 | 0.12) >>>>>> : H 4 : I 41 : P 20 :=20 >>>>>> snmp_counter:cpu_user_ticks(.1.3..0.0): 43094 -> >>>>>> buffer(): 3 (time P:1.12 | 0.12) >>>>>> : H 4 : I 41 : P 25 : snmp_counter:num_users(.1.3..5.0): 0 = -> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> buffer(): >>>>> >>>>>> 4 (time P:0.92 | 0.11) >>>>>> : H 4 : I 41 : P 30 :=20 >>>>>> snmp_counter:cpu_idle_ticks(.1.3..3.0): 6540643 >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> -> >>>>> >>>>>> buffer(): 5 (time P:1.08 | 0.11) >>>>>> : H 4 : I 41 : P 35 :=20 >>>>>> snmp_counter:tcp_established(.1.3..9.0): 68 -> >>>>>> buffer(): 6 (time P:0.84 | 0.11) >>>>>> : H 4 : I 41 : P 40 :=20 >>>>>> snmp_counter:cpu_system_ticks(.1.3..2.0): 79668 >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> -> >>>>> >>>>>> buffer(): 7 (time P:1.08 | 0.11) >>>>>> : H 4 : I 41 : P 45 : snmp_counter:tcp_active(.1.3..5.0):=20 >>>>>> 46088 -> >>>>>> buffer(): 8 (time P:0.71 | 0.11) >>>>>> : H 4 : I 41 : P 50 :=20 >>>>>> snmp_counter:load_average_1(.1.3..3.1): 0.06 -> >>>>>> buffer(): 9 (time P:0.84 | 0.13) >>>>>> : H 4 : I 41 : P 55 : snmp_counter:tcp_passive(.1.3..6.0):=20 >>>>>> 0 -> >>>>>> buffer(): 10 (time P:71.5 | 0.19) >>>>>> : H 4 : I 41 : P 60 :=20 >>>>>> snmp_counter:load_average_5(.1.3..3.2): 0.13 -> >>>>>> buffer(): 11 (time P:2.81 | 0.19) >>>>>> : H 4 : I 41 : P 70 :=20 >>>>>> snmp_counter:load_average_15(.1.3..3.3): 0.10 -> >>>>>> buffer(): 12 (time P:0.97 | 0.18) >>>>>> : H 4 : I 41 : P 75 : snmp_counter:mail_total(.1.3..1.1):=20 >>>>>> "1179" -> >>>>>> buffer(): 13 (time P:173.07 | 0.15) >>>>>> : H 4 : I 41 : P 80 : no_poller(): 0 -> rrd(*):=20 >>>>>> cpu_user_ticks:43094 - >>>>>> cpu_idle_ticks:6540643 - cpu_nice_ticks:1305 -=20 >>>>>> cpu_system_ticks:79668 - >>>>>> load_average_1:0.06 - load_average_5:0.13 - load_average_15:0.10 -= =20 >>>>>> num_users:0 >>>>>> - num_procs:261 - tcp_active:46088 - tcp_passive:0 -=20 >>>>>> tcp_established:68 (time >>>>>> P:0.19 | 117.95) >>>>>> : H 4 : I 41 : P LPD : last_poll_date(): 1094153380 -> >>>>>> db(last_poll_date): 1 (time P:0.21 | 2.29) >>>>>> : H 4 : Poller End, Total Time: 392.6 msec. >>>>>> >>>>>> The problem is at the no_poller() line, no value for mail_total,=20 >>>>>> but I don't >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> see >>>>> >>>>>> what's wrong in my new poller definition. >>>>>> >>>>>> Thanks for your help. >>>>>> Pierre >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------- >>>>> This SF.Net email is sponsored by BEA Weblogic Workshop >>>>> FREE Java Enterprise J2EE developer tools! >>>>> Get your free copy of BEA WebLogic Workshop 8.1 today. >>>>> http://ads.osdn.com/?ad_idP47&alloc_id=10808&op=3Dclick >>>>> _______________________________________________ >>>>> jffnms-users mailing list >>>>> jff...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/jffnms-users >>>> >>>> >>>> >>>> >>>> >>> >>> >>> ------------------------------------------------------- >>> This SF.Net email is sponsored by BEA Weblogic Workshop >>> FREE Java Enterprise J2EE developer tools! >>> Get your free copy of BEA WebLogic Workshop 8.1 today. >>> http://ads.osdn.com/?ad_id=3D5047&alloc_id=3D10808&op=3Dclick >>> _______________________________________________ >>> jffnms-users mailing list >>> jff...@li... >>> https://lists.sourceforge.net/lists/listinfo/jffnms-users >> >> >> --=20 =3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D= -=3D-=3D-=3D-=3D-=3D Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: <va...@cs...> - 2004-09-09 14:44:17
|
I have a problem where the disk SLA always reports 0. In the logs, the rrd_analyzer reports correctly the storage_block_count but zero for storage_block_size and storage_used_blocks. The poller and RRD graphs do show the correct values for the filesystem. Any pointers on how to debug this? Here is a sample from rrd_analyzer; 10:16:00 I82 : Start: 2004-09-09 09:40:00 Stop: 2004-09-09 10:10:00 Measures: 7 10:16:00 I82 : storage_block_size(0) storage_block_count(34812772) storage_used_blocks(0) 10:16:00 I82 : -------------------------------------------------------------------------= ----------------- 10:16:00 I82 : sla : Cond0: ( 0 > ((34812772*20)/100)) =3D 0 -FALSE- Used > 20%: 0 % 10:16:00 I82 : sla : Final Eval: False. JFFNMS v0.7.8 PHP v4.3.8 Net-SNMP 5.1.2(sever), 5.0.8(client) TIA - Jeff |
From: Javier S. <ja...@jf...> - 2004-09-09 15:01:56
|
Hi Jeff, This is odd.. For looking at the output I should say that the poller is not getting the correct values here. But you say the graph of this same interfaces looks ok. Can you show me the graph for interface 82 and the poller output for it too? Thanks, Javier va...@cs... wrote: > I have a problem where the disk SLA always reports 0. In the logs, the > rrd_analyzer reports correctly the storage_block_count but zero for > storage_block_size and storage_used_blocks. The poller and RRD graphs do > show the correct values for the filesystem. > > Any pointers on how to debug this? > > Here is a sample from rrd_analyzer; > > 10:16:00 I82 : Start: 2004-09-09 09:40:00 Stop: 2004-09-09 10:10:00 > Measures: 7 > 10:16:00 I82 : storage_block_size(0) storage_block_count(34812772) > storage_used_blocks(0) > 10:16:00 I82 : > ------------------------------------------------------------------------------------------ > 10:16:00 I82 : sla : Cond0: ( 0 > ((34812772*20)/100)) = 0 > -FALSE- Used > 20%: 0 % > 10:16:00 I82 : sla : Final Eval: False. > > JFFNMS v0.7.8 > PHP v4.3.8 > Net-SNMP 5.1.2(sever), 5.0.8(client) > > TIA - Jeff > > > ------------------------------------------------------- > This SF.Net email is sponsored by BEA Weblogic Workshop > FREE Java Enterprise J2EE developer tools! > Get your free copy of BEA WebLogic Workshop 8.1 today. > http://ads.osdn.com/?ad_idP47&alloc_id808&op=click > _______________________________________________ > jffnms-users mailing list > jff...@li... > https://lists.sourceforge.net/lists/listinfo/jffnms-users -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: Javier S. <ja...@jf...> - 2004-09-09 18:41:12
|
Hi Jeff, You are right, everything seems fine. Please add this debugs: var_dump ($values); in line 33 of engine/analyzers/sla.inc.php And then show me the output of the rrd analyzer again. Javier va...@cs... wrote: > Javier, > > The poller output is as follows; > > 11:30:25 : H 8 : I 82 : P 10 : > snmp_counter:storage_block_size(.1.3...17): 1024 -> buffer(): 19 (time > P:1.32 | 0.2) > 11:30:25 : H 8 : I 82 : P 20 : > snmp_counter:storage_block_count(.1.3...17): 34812772 -> buffer(): 37 > (time P:1.37 | 0.22) > 11:30:25 : H 8 : I 82 : P 30 : > snmp_counter:storage_used_blocks(.1.3...17): 4980071 -> buffer(): 54 (time > P:1.3 | 0.23) > 11:30:25 : H 8 : I 82 : P 60 : no_poller(): 0 -> rrd(*): > storage_block_size:1024 - storage_block_count:34812772 - > storage_used_blocks:4980071 (time P:0.17 | 27.92) > 11:30:26 : H 8 : I 82 : P LPD : last_poll_date(): 1094743825 -> > db(last_poll_date): 1 (time P:0.11 | 2.81) > > and the graph is attached > > > >>Hi Jeff, >> >>This is odd.. >>For looking at the output I should say that the poller is not getting the >>correct values here. >> >>But you say the graph of this same interfaces looks ok. >> >>Can you show me the graph for interface 82 and the poller output for it >>too? >> >>Thanks, >>Javier >> >>va...@cs... wrote: >> >>>I have a problem where the disk SLA always reports 0. In the logs, the >>>rrd_analyzer reports correctly the storage_block_count but zero for >>>storage_block_size and storage_used_blocks. The poller and RRD graphs do >>>show the correct values for the filesystem. >>> >>>Any pointers on how to debug this? >>> >>>Here is a sample from rrd_analyzer; >>> >>>10:16:00 I82 : Start: 2004-09-09 09:40:00 Stop: 2004-09-09 >>>10:10:00 >>> Measures: 7 >>>10:16:00 I82 : storage_block_size(0) storage_block_count(34812772) >>>storage_used_blocks(0) >>>10:16:00 I82 : >>>------------------------------------------------------------------------------------------ >>>10:16:00 I82 : sla : Cond0: ( 0 > ((34812772*20)/100)) = 0 >>>-FALSE- Used > 20%: 0 % >>>10:16:00 I82 : sla : Final Eval: False. >>> >>>JFFNMS v0.7.8 >>>PHP v4.3.8 >>>Net-SNMP 5.1.2(sever), 5.0.8(client) >>> >>>TIA - Jeff >>> >>> >>>------------------------------------------------------- >>>This SF.Net email is sponsored by BEA Weblogic Workshop >>>FREE Java Enterprise J2EE developer tools! >>>Get your free copy of BEA WebLogic Workshop 8.1 today. >>>http://ads.osdn.com/?ad_idP47&alloc_id808&op=click >>>_______________________________________________ >>>jffnms-users mailing list >>>jff...@li... >>>https://lists.sourceforge.net/lists/listinfo/jffnms-users >> >>-- >>=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= >>Javier Szyszlican, Project Leader, JFFNMS >>ja...@jf... >> >>I hope JFFNMS or I were helpful to you, if you >>can, please donate at http://jffnms.org/donate >> >> >> >>------------------------------------------------------- >>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 >>Project Admins to receive an Apple iPod Mini FREE for your judgement on >>who ports your project to Linux PPC the best. Sponsored by IBM. >>Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php >>_______________________________________________ >>jffnms-users mailing list >>jff...@li... >>https://lists.sourceforge.net/lists/listinfo/jffnms-users >> >> >> >> ------------------------------------------------------------------------ >> -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: <va...@cs...> - 2004-09-09 20:27:22
|
Javier, I checked all my rrd files and they all show a DS name of "data" but the rrdtool_fetch_ds function is passed "storage_block_size",=20 "storage_block_count", and "storage_used_blocks". Is the latter correct and all my rrd files were created with a generic DS of "data"? Jeff > Hi Jeff, > > You are right, everything seems fine. > > Please add this debugs: > > var_dump ($values); > > in line 33 of engine/analyzers/sla.inc.php > > And then show me the output of the rrd analyzer again. > > Javier > > va...@cs... wrote: >> Javier, >> >> The poller output is as follows; >> >> 11:30:25 : H 8 : I 82 : P 10 : >> snmp_counter:storage_block_size(.1.3...17): 1024 -> buffer(): 19 (time >> P:1.32 | 0.2) >> 11:30:25 : H 8 : I 82 : P 20 : >> snmp_counter:storage_block_count(.1.3...17): 34812772 -> buffer(): 37 >> (time P:1.37 | 0.22) >> 11:30:25 : H 8 : I 82 : P 30 : >> snmp_counter:storage_used_blocks(.1.3...17): 4980071 -> buffer(): 54 >> (time >> P:1.3 | 0.23) >> 11:30:25 : H 8 : I 82 : P 60 : no_poller(): 0 -> rrd(*): >> storage_block_size:1024 - storage_block_count:34812772 - >> storage_used_blocks:4980071 (time P:0.17 | 27.92) >> 11:30:26 : H 8 : I 82 : P LPD : last_poll_date(): 1094743825 -> >> db(last_poll_date): 1 (time P:0.11 | 2.81) >> >> and the graph is attached >> >> >> >>>Hi Jeff, >>> >>>This is odd.. >>>For looking at the output I should say that the poller is not getting >>> the >>>correct values here. >>> >>>But you say the graph of this same interfaces looks ok. >>> >>>Can you show me the graph for interface 82 and the poller output for i= t >>>too? >>> >>>Thanks, >>>Javier >>> >>>va...@cs... wrote: >>> >>>>I have a problem where the disk SLA always reports 0. In the logs, th= e >>>>rrd_analyzer reports correctly the storage_block_count but zero for >>>>storage_block_size and storage_used_blocks. The poller and RRD graphs >>>> do >>>>show the correct values for the filesystem. >>>> >>>>Any pointers on how to debug this? >>>> >>>>Here is a sample from rrd_analyzer; >>>> >>>>10:16:00 I82 : Start: 2004-09-09 09:40:00 Stop: 2004-09-09 >>>>10:10:00 >>>> Measures: 7 >>>>10:16:00 I82 : storage_block_size(0) storage_block_count(34812772) >>>>storage_used_blocks(0) >>>>10:16:00 I82 : >>>>---------------------------------------------------------------------= --------------------- >>>>10:16:00 I82 : sla : Cond0: ( 0 > ((34812772*20)/100)) =3D= 0 >>>>-FALSE- Used > 20%: 0 % >>>>10:16:00 I82 : sla : Final Eval: False. >>>> >>>>JFFNMS v0.7.8 >>>>PHP v4.3.8 >>>>Net-SNMP 5.1.2(sever), 5.0.8(client) >>>> >>>>TIA - Jeff >>>> >>>> >>>>------------------------------------------------------- >>>>This SF.Net email is sponsored by BEA Weblogic Workshop >>>>FREE Java Enterprise J2EE developer tools! >>>>Get your free copy of BEA WebLogic Workshop 8.1 today. >>>>http://ads.osdn.com/?ad_idP47&alloc_id=10808&op=3Dclick >>>>_______________________________________________ >>>>jffnms-users mailing list >>>>jff...@li... >>>>https://lists.sourceforge.net/lists/listinfo/jffnms-users >>> >>>-- >>>=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D= -=3D-=3D-=3D-=3D-=3D-=3D >>>Javier Szyszlican, Project Leader, JFFNMS >>>ja...@jf... >>> >>>I hope JFFNMS or I were helpful to you, if you >>>can, please donate at http://jffnms.org/donate >>> >>> >>> >>>------------------------------------------------------- >>>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 >>>Project Admins to receive an Apple iPod Mini FREE for your judgement o= n >>>who ports your project to Linux PPC the best. Sponsored by IBM. >>>Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php >>>_______________________________________________ >>>jffnms-users mailing list >>>jff...@li... >>>https://lists.sourceforge.net/lists/listinfo/jffnms-users >>> >>> >>> >>> ---------------------------------------------------------------------= --- >>> > > -- > =3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D= -=3D-=3D-=3D-=3D-=3D-=3D > Javier Szyszlican, Project Leader, JFFNMS > ja...@jf... > > I hope JFFNMS or I were helpful to you, if you > can, please donate at http://jffnms.org/donate > > > > ------------------------------------------------------- > This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 > Project Admins to receive an Apple iPod Mini FREE for your judgement on > who ports your project to Linux PPC the best. Sponsored by IBM. > Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php > _______________________________________________ > jffnms-users mailing list > jff...@li... > https://lists.sourceforge.net/lists/listinfo/jffnms-users > |
From: Javier S. <ja...@jf...> - 2004-09-09 20:28:46
|
Jeff, Thats how JFFNMS works. The real name is only stored in the DB with a reference to the file it is. So, all your RRD Files will have a DS named 'data'. Javier va...@cs... wrote: > Javier, > > I checked all my rrd files and they all show a DS name of "data" but the > rrdtool_fetch_ds function is passed "storage_block_size", > "storage_block_count", and "storage_used_blocks". > > Is the latter correct and all my rrd files were created with a generic DS > of "data"? > > Jeff > > >>Hi Jeff, >> >>You are right, everything seems fine. >> >>Please add this debugs: >> >>var_dump ($values); >> >>in line 33 of engine/analyzers/sla.inc.php >> >>And then show me the output of the rrd analyzer again. >> >>Javier >> >>va...@cs... wrote: >> >>>Javier, >>> >>>The poller output is as follows; >>> >>>11:30:25 : H 8 : I 82 : P 10 : >>>snmp_counter:storage_block_size(.1.3...17): 1024 -> buffer(): 19 (time >>>P:1.32 | 0.2) >>>11:30:25 : H 8 : I 82 : P 20 : >>>snmp_counter:storage_block_count(.1.3...17): 34812772 -> buffer(): 37 >>>(time P:1.37 | 0.22) >>>11:30:25 : H 8 : I 82 : P 30 : >>>snmp_counter:storage_used_blocks(.1.3...17): 4980071 -> buffer(): 54 >>>(time >>>P:1.3 | 0.23) >>>11:30:25 : H 8 : I 82 : P 60 : no_poller(): 0 -> rrd(*): >>>storage_block_size:1024 - storage_block_count:34812772 - >>>storage_used_blocks:4980071 (time P:0.17 | 27.92) >>>11:30:26 : H 8 : I 82 : P LPD : last_poll_date(): 1094743825 -> >>>db(last_poll_date): 1 (time P:0.11 | 2.81) >>> >>>and the graph is attached >>> >>> >>> >>> >>>>Hi Jeff, >>>> >>>>This is odd.. >>>>For looking at the output I should say that the poller is not getting >>>>the >>>>correct values here. >>>> >>>>But you say the graph of this same interfaces looks ok. >>>> >>>>Can you show me the graph for interface 82 and the poller output for it >>>>too? >>>> >>>>Thanks, >>>>Javier >>>> >>>>va...@cs... wrote: >>>> >>>> >>>>>I have a problem where the disk SLA always reports 0. In the logs, the >>>>>rrd_analyzer reports correctly the storage_block_count but zero for >>>>>storage_block_size and storage_used_blocks. The poller and RRD graphs >>>>>do >>>>>show the correct values for the filesystem. >>>>> >>>>>Any pointers on how to debug this? >>>>> >>>>>Here is a sample from rrd_analyzer; >>>>> >>>>>10:16:00 I82 : Start: 2004-09-09 09:40:00 Stop: 2004-09-09 >>>>>10:10:00 >>>>> Measures: 7 >>>>>10:16:00 I82 : storage_block_size(0) storage_block_count(34812772) >>>>>storage_used_blocks(0) >>>>>10:16:00 I82 : >>>>>------------------------------------------------------------------------------------------ >>>>>10:16:00 I82 : sla : Cond0: ( 0 > ((34812772*20)/100)) = 0 >>>>>-FALSE- Used > 20%: 0 % >>>>>10:16:00 I82 : sla : Final Eval: False. >>>>> >>>>>JFFNMS v0.7.8 >>>>>PHP v4.3.8 >>>>>Net-SNMP 5.1.2(sever), 5.0.8(client) >>>>> >>>>>TIA - Jeff >>>>> >>>>> >>>>>------------------------------------------------------- >>>>>This SF.Net email is sponsored by BEA Weblogic Workshop >>>>>FREE Java Enterprise J2EE developer tools! >>>>>Get your free copy of BEA WebLogic Workshop 8.1 today. >>>>>http://ads.osdn.com/?ad_idP47&alloc_id808&op=click >>>>>_______________________________________________ >>>>>jffnms-users mailing list >>>>>jff...@li... >>>>>https://lists.sourceforge.net/lists/listinfo/jffnms-users >>>> >>>>-- >>>>=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= >>>>Javier Szyszlican, Project Leader, JFFNMS >>>>ja...@jf... >>>> >>>>I hope JFFNMS or I were helpful to you, if you >>>>can, please donate at http://jffnms.org/donate >>>> >>>> >>>> >>>>------------------------------------------------------- >>>>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 >>>>Project Admins to receive an Apple iPod Mini FREE for your judgement on >>>>who ports your project to Linux PPC the best. Sponsored by IBM. >>>>Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php >>>>_______________________________________________ >>>>jffnms-users mailing list >>>>jff...@li... >>>>https://lists.sourceforge.net/lists/listinfo/jffnms-users >>>> >>>> >>>> >>>>------------------------------------------------------------------------ >>>> >> >>-- >>=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= >>Javier Szyszlican, Project Leader, JFFNMS >>ja...@jf... >> >>I hope JFFNMS or I were helpful to you, if you >>can, please donate at http://jffnms.org/donate >> >> >> >>------------------------------------------------------- >>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 >>Project Admins to receive an Apple iPod Mini FREE for your judgement on >>who ports your project to Linux PPC the best. Sponsored by IBM. >>Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php >>_______________________________________________ >>jffnms-users mailing list >>jff...@li... >>https://lists.sourceforge.net/lists/listinfo/jffnms-users > >> > -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |
From: Brad E. <ben...@gm...> - 2005-02-22 15:15:50
|
We continue to get Warmstart reports on our APC Symmetra. This didn't start until we started monitoring with JFFNMS 0.8.0. I'm assuming it has something to do with the APC pollers introduced in the new version. Any ideas what I can do to solve this or why it may be doing this? Brad Endsley Information Technology TechSpray, L.P. |
From: Javier S. <ja...@jf...> - 2005-02-22 16:36:55
|
Hi Brad, You have to check if this messages come from syslog or are generated by p= olling=20 some OID. If they are not from syslog, I'll need you to get me a snmpwalk of=20 .1.3.6.1.4.1.318 so I can debug it. Also, send me the exact event you see in the events screen. Thanks, Javier Brad Endsley wrote: > We continue to get Warmstart reports on our APC Symmetra. This didn't=20 > start until we started monitoring with JFFNMS 0.8.0. I=92m assuming it= =20 > has something to do with the APC pollers introduced in the new version. >=20 > =20 >=20 > Any ideas what I can do to solve this or why it may be doing this? >=20 > =20 >=20 > =20 >=20 > Brad Endsley > Information Technology > TechSpray, L.P. >=20 --=20 =3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D= -=3D-=3D-=3D-=3D-=3D Javier Szyszlican, Project Leader, JFFNMS ja...@jf... I hope JFFNMS or I were helpful to you, if you can, please donate at http://jffnms.org/donate |