- Nav-users - lister.unit.no

postgres and "nonstandard use of \\"
by Ingeborg.Hellemo＠cc.uit.no 10 Okt '07

10 Okt '07

My postgres 8.2 is whining about "WARNING: nonstandard use of \\ in a string literal at character NN" "HINT: Use the escape string syntax for backslashes, e.g., E'\\'." I think I have located the source of the warnings to be these lines in /usr/local/nav/doc/sql/navprofiles.sql (NAV 3.3.0) INSERT INTO AccountGroupPrivilege (accountgroupid, privilegeid, target) VALUES (2, 2, '^/toolbox\\b'); INSERT INTO AccountGroupPrivilege (accountgroupid, privilegeid, target) VALUES (2, 2, '^/index(.py)?/(index|login|logout|userinfo|passwd)\\b'); INSERT INTO AccountGroupPrivilege (accountgroupid, privilegeid, target) VALUES (2, 2, '^/messages/(active|historic|planned|view|rss)\\b'); INSERT INTO AccountGroupPrivilege (accountgroupid, privilegeid, target) VALUES (2, 2, '^/maintenance/(calendar|active|historic|planned|view)\\b'); INSERT INTO AccountGroupPrivilege (accountgroupid, privilegeid, target) VALUES (2, 2, '^/vlanPlot\\b'); INSERT INTO AccountGroupPrivilege (accountgroupid, privilegeid, target) VALUES (2, 2, '^/vPServer/servlet/vPServer\\b'); Are the lines supposed to look like this? --Ingeborg -- Ingeborg ?strem Hellemo -- ingeborg(a)cc.uit.no (Univ. of Troms?, Norway)

1 0

cricket on new install
by ctc＠ctcweb.net 09 Okt '07

09 Okt '07

New Debian Etch install using package. Received the following error. goozzy:/etc/cricket/config# /usr/lib/nav/makecricketconfig.pl No path to cricket-conf.pl specified, searching default paths... Searching in /home/navcron/cricket/cricket. Searching in /etc/cricket. Found cricket-conf.pl in /etc/cricket. Failed to open compiled config tree from gConfigRoot/config.db: No such file or directory If this is the first time you run makecricketconfig, try compile cricket (/etc/cricket/cricket/compile) goozzy:/etc/cricket/config# Thanks for any help you can provide. /etc/cricket/cricket-config.pl looks like this: $gCricketHome = "/usr/share/cricket"; $gInstallRoot = "/usr/share/cricket"; $gConfigRoot = "/etc/nav/cricket-config"; $gLogDir = "/var/log/cricket"; # Whether to optimize config tree access for the collector by # copying the database into memory on startup. # Switch off if your collector becomes too big in run time. # Most sites will probably want slurp, this is the default. #$gDbAccess = "slurp"; # Slurp the database into memory #$gDbAccess = "lookup"; # Use normal DB lookups # Where cricket caches generated images. $gCacheDir = "/var/cache/cricket"; # It is strongly recommended that you *not* set a global logLevel. # Use the CRICKET_LOG_LEVEL environment variable or the # -logLevel option instead. #$gLogLevel = "info"; per the README Thanks for any help. I saw an old README that showed changing CricketHome not ConfigRoot - is that the issue? Caryl A. Whitlatch Network Administrator HYPERLINK "http://www.ctcweb.net/"CTC Telecom, Inc. Plurima ignoro, sed ignorantiam meam non ignoro No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.488 / Virus Database: 269.14.4/1057 - Release Date: 10/8/2007 9:04 AM >From cwhitlatch at ctctele.com Tue Oct 9 11:24:00 2007 From: cwhitlatch at ctctele.com (Mr. Caryl Whitlatch) Date: Tue Oct 9 18:23:38 2007 Subject: [Nav-users] cricket answer Message-ID: <001101c80a90$ccf62170$2632a8c0@cwhitlatch> THANKS, Werner The README.Debian has been changed for the next version of the package. The correct procedure is as follows after changing the /etc/cricket/cricket-conf.pl (your looks correct): 'sudo -u navcron cricket-compile' 'sudo -u navcron /usr/lib/nav/makecricketconfig.pl' Good luck and thanks for using the Debian package! - Werner Caryl A. Whitlatch Network Administrator HYPERLINK "http://www.ctcweb.net/"CTC Telecom, Inc. Plurima ignoro, sed ignorantiam meam non ignoro No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.488 / Virus Database: 269.14.6/1059 - Release Date: 10/9/2007 8:44 AM No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.488 / Virus Database: 269.14.6/1059 - Release Date: 10/9/2007 8:44 AM >From prieheck at iwu.edu Tue Oct 9 17:24:31 2007 From: prieheck at iwu.edu (Pat Riehecky) Date: Tue Oct 9 23:24:57 2007 Subject: [Nav-users] Java error Message-ID: <1191965071.15629.17.camel(a)thales.iwu.edu> I am getting the following error in the networkDiscovery log: Warning: -Xmx256m not understood. Ignoring. Warning: -jar not understood. Ignoring. Exception in thread "main" java.lang.NoClassDefFoundError: networkDiscovery.jar at gnu.gcj.runtime.FirstThread.run() (/usr/lib/libgcj.so.5.0.0) at _Jv_ThreadRun(java.lang.Thread) (/usr/lib/libgcj.so.5.0.0) at _Jv_RunMain(java.lang.Class, byte const, int, byte const, boolean) (/usr/lib/libgcj.so.5.0.0) at __gcj_personality_v0 (/usr/local/nav/lib/java/networkDiscovery/java.version=1.4.2) at __libc_start_main (/lib/tls/libc-2.3.4.so) at _Jv_RegisterClasses (/usr/local/nav/lib/java/networkDiscovery/java.version=1.4.2) How do I fix this... I am not a java person..... NAV 3.2.2, RHEL4 Pat

1 0

NAV 3.3.0 RC1 + Unsupported SNMP protocol version
by morten.brekkevold＠uninett.no 05 Okt '07

05 Okt '07

Erik Weber wrote, On 01-10-2007 14:08: >> Question for you: Does this happen when >> you insert single boxes using the Add form, or does it happen only >> when bulk >> importing? Or does it happen on both? >> > I've only tried to add single boxes, and the error message occur pretty > fast (less than 2 seconds), so that timeout value is pretty low, hehe. > > I can try doing a bulk import if you think it matters. No, that won't be necessary. I've just tested it some more and found that this only happens if no response is received, not even an ICMP port-not-reachable message. If the NAV server receives such an ICMP message from the target or the router, it will instead display "Connection refused". I guess this indicates that the SNMP UDP request packet is lost on the way to the target, either dropped by some firewall or the box itself, or the return packet is dropped in a similar fashion. I will add an issue for this in the bugtracker, but for now, the error message can be intepreted as a timeout error. -- mvh Morten Brekkevold UNINETT >From morten.brekkevold at uninett.no Fri Oct 5 12:33:01 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Fri Oct 5 11:33:52 2007 Subject: [Nav-users] Announcement: NAV 3.3.0 RC1 released In-Reply-To: <4700E4AD.3060709(a)gmail.com> References: <46EFCB84.70504(a)uninett.no> <46F8AC23.8090409(a)gmail.com> <4700B683.4050103(a)uninett.no> <4700E4AD.3060709(a)gmail.com> Message-ID: <470604CD.9050004(a)uninett.no> Erik Weber wrote, On 01-10-2007 14:14: >> I suspect you are thinking of process monitoring through SNMP? NAV >> already supports external monitoring of the most common TCP/IP reachable >> services, but does not currently support monitoring internal server >> processes through SNMP. > That's correct. We're looking for a new tool to do our surveillance and > this includes a few servers with some critical services as well as com. > equipment. Exactly, well, then Ole Martin is the man to talk to :) -- mvh Morten Brekkevold UNINETT >From morten.brekkevold at uninett.no Fri Oct 5 12:37:16 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Fri Oct 5 11:37:19 2007 Subject: [Nav-users] Re: java.lang.NoClassDefFoundError: no.ntnu.nav.SimpleSnmp.SimpleSnmp In-Reply-To: <200710021540.52952.marco.barbieri(a)unimore.it> References: <47010BE8.1000808(a)uninett.no> <200710021540.52952.marco.barbieri(a)unimore.it> Message-ID: <470605CC.8070903(a)uninett.no> Marco Barbieri wrote, On 02-10-2007 15:40: > hi, all > maybe it's just me but i'm still getting some errors fro getBoksMacs.sh, i > think that i've cleaned/reinstalled quite everything needed but, you know... > > anyway here comes the error from a Centos 5 install recived by mail: > > Today 15:26:16 [...] > Exception in thread "QueryBoks-20" Exception in thread "QueryBoks-21" > java.lang.NoClassDefFoundError: no.ntnu.nav.SimpleSnmp.SimpleSnmp > at java.lang.Class.initializeClass(libgcj.so.7rh) > at QueryBoks.run(QueryBoks.java:200) > Caused by: java.lang.ClassNotFoundException: snmp.SNMPVarBindList not found in > gnu.gcj.runtime.ExtensionClassLoader{urls=[file:/usr/local/nav/lib/java/Database.jar,file:/usr/local/nav/lib/java/Event.jar,file:/usr/local/nav/lib/java/SimpleSnmp.jar,file:/usr/local/nav/lib/java/postgresql-jdbc.jar,file:/usr/local/nav/lib/java/Logger.jar,file:/usr/local/nav/lib/java/ConfigParser.jar,file:/usr/local/nav/lib/java/NetboxInfo.jar,file:/usr/local/nav/lib/java/Util.jar], > parent=null} hm, check the contents of /usr/local/nav/lib/java/SimpleSnmp.jar: jar -tf /usr/local/nav/lib/java/SimpleSnmp.jar If if contains no/ntnu/nav/SimpleSnmp/SimpleSnmp.class, then something is wrong with your Java runtine environment. Doesn't look like you're using Sun's JVM, which I would recommend in any case. > hope it's just me ;-) So do I :) -- mvh Morten Brekkevold UNINETT >From NPabon at archstonesmith.com Mon Oct 8 12:09:51 2007 From: NPabon at archstonesmith.com (Pabon, Nestor) Date: Mon Oct 8 19:10:01 2007 Subject: [Nav-users] Could not connect to servlet Message-ID: <58B7A531794E6A49B3BAD57E333225440249A320(a)engexc11.archstonesmith.com> When attempting selecting the Network Explorer I get the following: Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/nav/web/pyproxy/proxy.py", line 61, in handler user=user, session=req.session.id) File "/usr/lib/python2.4/site-packages/nav/web/pyproxy/proxy.py", line 124, in doRequest request.request(method, path, query, headers) File "httplib.py", line 804, in request self._send_request(method, url, body, headers) File "httplib.py", line 827, in _send_request self.endheaders() File "httplib.py", line 798, in endheaders self._send_output() File "httplib.py", line 679, in _send_output self.send(msg) File "httplib.py", line 646, in send self.connect() File "httplib.py", line 630, in connect raise socket.error, msg error: (111, 'Connection refused') This is a packaged installation of NAV 3.2.2. on a Debian Etch 4.0. Thank you in advance for your assistance and guidance Nestor Pabon Sr. Network Engineer

1 0

Traffic Map will not display
by rkillick＠niagara.com 03 Okt '07

03 Okt '07

I have recently installed Nav 3.2.2 onto a CentOS 4.5 workstation using the available installation notes. I have been able to seed the database (6 xGSW + 43 x SW) and have tested most of the tool box tools without much problem with the exception of the Traffic Map. The Traffic Map will not display and simply shows an empty box where the map should be, and a lone remark in the status bar at the bottom left "Applet vlanPlot notinted". Note: The "IP address scope - graphical view" works fine. I have placed 6 x GSW (C3750s) into the '_Top' map container. Any ideas as to where to begin looking for the problem would be appreciated. Bob Killick

1 0

Announcement: NAV 3.3.0 RC1 released
by morten.brekkevold＠uninett.no 01 Okt '07

01 Okt '07

Gheorghe Limboi wrote, On 23-09-2007 00:13: > Hi, > I got this error when trying to make from sources. > > Building getDeviceData ... > (cd getDeviceData && /usr/local/ant/bin/ant ) && touch getDeviceData > Buildfile: build.xml > init: > [mkdir] Created dir: /home/ghita/proiect/nav-3.3.0_rc1/src/getDeviceData/build > compile: > [javac] Compiling 15 source files to /home/ghita/proiect/nav-3.3.0_rc1/src/getDeviceData/build > [javac] /home/ghita/proiect/nav-3.3.0_rc1/src/getDeviceData/src/OidTester.java:41: cannot find symbol > [javac] symbol : method getSnmpVersion() > [javac] location: class no.ntnu.nav.SimpleSnmp.SimpleSnmp > [javac] Log.d("OID_TESTER", "DO_TEST", "Using SNMP version " + sSnmp.getSnmpVersion() + " for " + nb.getSysname()); > [javac] ^ > [javac] /home/ghita/proiect/nav-3.3.0_rc1/src/getDeviceData/src/OidTester.java:252: cannot find symbol > [javac] symbol : method getNext(int,boolean,boolean,boolean) > [javac] location: class no.ntnu.nav.SimpleSnmp.SimpleSnmp > [javac] l = sSnmp.getNext(getCnt, snmpoid.getDecodehex(), true, false); > [javac] ^ > [javac] 2 errors > BUILD FAILED > > > I notice some diff from OidTester.java in 3.3.0 and 3.2.2 . After replacing > OidTester.java with the 3.2.2 one I was able to compile without errors. > > echo $CLASSPATH > /usr/lib/jdk1.5.0_04/jre/lib/ext/snmp.jar:/usr/lib/jdk1.5.0_04/jre/lib/ext/servlet-api.jar I'm not sure why you are getting these errors, a build from scratch on a clean server works fine here. The two methods your compiler is unable to find were added to SimpleSnmp.java after NAV 3.2.2, and are present in the 3.3 release candidate. It's as if you are compiling a new version of getDeviceData against an old version of SimpleSnmp. Of course, when you replace OidTester.java with a version as old as the SimpleSnmp.java without the two new symbols, you will have success in compiling, but you will experience other problems during runtime (such as the Drexel SNMP memory leak / inifinite loop problem). Are you sure /home/ghita/proiect/nav-3.3.0_rc1/ does not contain remnants of old NAV versions? -- mvh Morten Brekkevold UNINETT >From morten.brekkevold at uninett.no Mon Oct 1 11:57:39 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Mon Oct 1 10:57:42 2007 Subject: [Nav-users] Announcement: NAV 3.3.0 RC1 released In-Reply-To: <46F8AC23.8090409(a)gmail.com> References: <46EFCB84.70504(a)uninett.no> <46F8AC23.8090409(a)gmail.com> Message-ID: <4700B683.4050103(a)uninett.no> Erik Weber wrote, On 25-09-2007 08:35: > Glad to see a new release, does this version support surveillance of > services on snmp enabled servers? I suspect you are thinking of process monitoring through SNMP? NAV already supports external monitoring of the most common TCP/IP reachable services, but does not currently support monitoring internal server processes through SNMP. > If not, do you have any idea of which release/when that could be > implemented? I'm currently looking for a new network management system, > but I want something that can monitor all my systems. Due to limited resources, the core NAV developer team is not focusing on new service monitoring functionality, but on the core network monitoring functionality. The University of Troms?, however, are interested in using, maintaining and developing further the service monitoring aspects of NAV. You should maybe have a chat with Ole Martin Bj?rndalen of the University of Troms?, whom I've Cc:ed in this e-mail. He has written a NAV subsystem which does exactly what I think you are after: It monitors server processes through SNMP. -- mvh Morten Brekkevold UNINETT >From morten.brekkevold at uninett.no Mon Oct 1 12:23:49 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Mon Oct 1 11:23:57 2007 Subject: [Nav-users] NAV 3.3.0 RC1 + Unsupported SNMP protocol version In-Reply-To: <46FB9C0B.7020109(a)gmail.com> References: <46FB9C0B.7020109(a)gmail.com> Message-ID: <4700BCA5.20406(a)uninett.no> Erik Weber wrote, On 27-09-2007 14:03: > When I try to add boxes I often get the following error message: > nav.Snmp.UnsupportedSnmpVersionError: Unsupported SNMP protocol version: 0 > > Anyone have an idea of how to fix it ? Debian Etch with the latest > package from Werner. A quick look at the code reveals a bizarre procedure; EditDB attempts to find out which SNMP version the box supports. It first attempts version 2c, if this times out it tries version 1, if this also times out it claims the SNMP version is 0. Apparently this value is used in later requests, which, of course, fails miserably. So, to make it short: It seems that what really happens when you get this message is that SNMP communication with the box timed out. I need to inspect the code a bit closer before I can tell when this actually occurs and figure out how to fix it. Question for you: Does this happen when you insert single boxes using the Add form, or does it happen only when bulk importing? Or does it happen on both? I have filed this as a bug here: http://sourceforge.net/support/tracker.php?aid=1805509 -- mvh Morten Brekkevold UNINETT >From twiztar+nav at gmail.com Mon Oct 1 15:08:28 2007 From: twiztar+nav at gmail.com (Erik Weber) Date: Mon Oct 1 14:08:45 2007 Subject: [Nav-users] NAV 3.3.0 RC1 + Unsupported SNMP protocol version In-Reply-To: <4700BCA5.20406(a)uninett.no> References: <46FB9C0B.7020109(a)gmail.com> <4700BCA5.20406(a)uninett.no> Message-ID: <4700E33C.4020002(a)gmail.com> Morten Brekkevold wrote: > Erik Weber wrote, On 27-09-2007 14:03: > >> When I try to add boxes I often get the following error message: >> nav.Snmp.UnsupportedSnmpVersionError: Unsupported SNMP protocol version: 0 >> >> Anyone have an idea of how to fix it ? Debian Etch with the latest >> package from Werner. >> > > Question for you: Does this happen when > you insert single boxes using the Add form, or does it happen only when bulk > importing? Or does it happen on both? > I've only tried to add single boxes, and the error message occur pretty fast (less than 2 seconds), so that timeout value is pretty low, hehe. I can try doing a bulk import if you think it matters. -- Erik Weber >From twiztar at gmail.com Mon Oct 1 15:14:37 2007 From: twiztar at gmail.com (Erik Weber) Date: Mon Oct 1 14:15:49 2007 Subject: [Nav-users] Announcement: NAV 3.3.0 RC1 released In-Reply-To: <4700B683.4050103(a)uninett.no> References: <46EFCB84.70504(a)uninett.no> <46F8AC23.8090409(a)gmail.com> <4700B683.4050103(a)uninett.no> Message-ID: <4700E4AD.3060709(a)gmail.com> Morten Brekkevold wrote: > Erik Weber wrote, On 25-09-2007 08:35: > > >> Glad to see a new release, does this version support surveillance of >> services on snmp enabled servers? >> > > I suspect you are thinking of process monitoring through SNMP? NAV already > supports external monitoring of the most common TCP/IP reachable services, but > does not currently support monitoring internal server processes through SNMP. > > That's correct. We're looking for a new tool to do our surveillance and this includes a few servers with some critical services as well as com. equipment. -- Erik Weber >From morten.brekkevold at uninett.no Mon Oct 1 18:02:00 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Mon Oct 1 17:02:10 2007 Subject: [Nav-users] Announcement: NAV 3.3.0 released Message-ID: <47010BE8.1000808(a)uninett.no> NAV 3.3.0 is now available for download at SourceForge, see http://sourceforge.net/projects/nav :-) This is mainly a feature release, but many bugs have also been fixed. Feedback from users and our own testing efforts revealed a few small glitches in the release candidate, and these have been fixed in this final release version. The Debian package maintainer, Morten Werner Olsen, is working on a Debian Etch package of NAV 3.3, which will be released shortly. Happy NAVing, everyone! -- mvh Morten Brekkevold UNINETT

1 0

Announcement: NAV 3.3.0 RC1 released
by morten.brekkevold＠uninett.no 18 Sep '07

18 Sep '07

A release candidate of NAV 3.3.0 is now available for download at SourceForge, see http://sourceforge.net/projects/nav :-) This is mainly a feature release, but many bugs have also been fixed. We kindly ask you to give us feedback, both positive and negative, on the mailing list if you decide to test the release candidate. We will wait about 10 days to see what issues may arise concering this candidate, before we release the final 3.3.0 version. The Debian package maintainer, Morten Werner Olsen, is currently in the final stages of testing a Debian Etch package of NAV 3.2.2, and will proceed to package NAV 3.3.0 when this work is finished. -- mvh Morten Brekkevold UNINETT

1 0

Re: rrdtool and other problems on Suse 10.2
by morten.brekkevold＠uninett.no 07 Sep '07

07 Sep '07

Mels wrote, On 09/06/2007 09:01 PM: > Hi all, Hi Mels. > I started this week with a clean system, installing nav on Suse 10.2 > > There was a problem with connecting to a remote database, for the time i > use the local database, witch is OK for nav. Ok, if you need some tips on connecting to a remote database, please don't hesitate to ask :) > I get 3 problems: > 1 I installed py-rrdtool-1.0b1, with the installation off > py-rrdtool-0.2.1 I get a error: > building 'rrdtool._rrdtool' extension > gcc -pthread -fno-strict-aliasing -DNDEBUG -O2 -march=i586 -mtune=i686 > -fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -fPIC > -I/usr/local/include -I/usr/include/python2.5 -c src/_rrdtoolmodule.c -o > build/temp.linux-i686-2.5/src/_rrdtoolmodule.o > src/_rrdtoolmodule.c: In function ?PyRRD_graph?: > src/_rrdtoolmodule.c:239: error: too few arguments to function ?rrd_graph? > error: command 'gcc' failed with exit status 1 I did not recognize the version numbers you mentioned, and checking up on things seems to reveal that there is some confusion about which python-rrdtool library to use. Apparently there are two very similar libraries in existance: One is distributed alongside rrdtool itself, and one is available at http://sourceforge.net/projects/py-rrdtool . The one currently in use by us on our dev server is the python-rrd library that comes with rrdtool itself; in Debian this package is called python-rrd. The package also follows the same version numbering scheme as rrdtool, which means that the latest version of it is >= 1.2. Unless Suse 10.2 prepackages this alongside its regular rrdtool package, you'll need to get hold of the rrdtool source code instead. This would also be the best way to make sure the python library version matches the rrdtool library version. > 2 and 3 pping and servicemon failed to start, I tried a nav start pping > servicemon a couple of time, also a start separate. > but they whon't start. I changed DEBUG_LEVEL = 6 in nav.conf, but can't > find any reason in the log files. The DEBUG_LEVEL in nav.conf only sets debug levels for the Java subsystems (i.e. getDeviceData, eventEngine, getBoksMacs, networkDiscovery). The pping and servicemon processes have their own debug level settings in their respective config files. That said, pping and servicemon share much of the same code base, and they are both known to have startup difficulties. We have yet to investigate what causes the problem, but usually, when one experiences startup problems with these two, it helps to retry it several times. -- mvh Morten Brekkevold UNINETT >From mels at mels-kooijman.nl Fri Sep 7 21:12:33 2007 From: mels at mels-kooijman.nl (Mels) Date: Fri Sep 7 20:12:23 2007 Subject: [Nav-users] Re: rrdtool and other problems on Suse 10.2 In-Reply-To: <46E13E6C.3020600(a)uninett.no> References: <200709031022.52942.barbieri.marco(a)unimore.it> <46DFC491.4070702(a)uninett.no> <46E04EA1.3010906(a)mels-kooijman.nl> <46E13E6C.3020600(a)uninett.no> Message-ID: <46E19491.2050804(a)mels-kooijman.nl> Thanks for your reply. I removed the rrdtool, wich was from the suse media. I compiled RRDtool 1.2.15 (the suse version was the same) and install it. I recompiled psycopg-1.1.21, the cronjob produced a error. After this restartapache,tomcat,postgresql and nav Even pping and servicemon started the first time. No errormail from the cronjobs, everythings looks fine, til i login in nav: error File "/usr/local/nav/lib/python/nav/web/state.py", line 250, in __new__ raise e Maybe i have to rebuild more modules? Mels Morten Brekkevold wrote: > Mels wrote, On 09/06/2007 09:01 PM: > > >> Hi all, >> > > Hi Mels. > > >> I started this week with a clean system, installing nav on Suse 10.2 >> >> There was a problem with connecting to a remote database, for the time i >> use the local database, witch is OK for nav. >> > > Ok, if you need some tips on connecting to a remote database, please don't > hesitate to ask :) > > >> I get 3 problems: >> 1 I installed py-rrdtool-1.0b1, with the installation off >> py-rrdtool-0.2.1 I get a error: >> building 'rrdtool._rrdtool' extension >> gcc -pthread -fno-strict-aliasing -DNDEBUG -O2 -march=i586 -mtune=i686 >> -fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -fPIC >> -I/usr/local/include -I/usr/include/python2.5 -c src/_rrdtoolmodule.c -o >> build/temp.linux-i686-2.5/src/_rrdtoolmodule.o >> src/_rrdtoolmodule.c: In function ?PyRRD_graph?: >> src/_rrdtoolmodule.c:239: error: too few arguments to function ?rrd_graph? >> error: command 'gcc' failed with exit status 1 >> > > I did not recognize the version numbers you mentioned, and checking up on > things seems to reveal that there is some confusion about which python-rrdtool > library to use. Apparently there are two very similar libraries in existance: > One is distributed alongside rrdtool itself, and one is available at > http://sourceforge.net/projects/py-rrdtool . > > The one currently in use by us on our dev server is the python-rrd library > that comes with rrdtool itself; in Debian this package is called python-rrd. > The package also follows the same version numbering scheme as rrdtool, which > means that the latest version of it is >= 1.2. Unless Suse 10.2 prepackages > this alongside its regular rrdtool package, you'll need to get hold of the > rrdtool source code instead. This would also be the best way to make sure the > python library version matches the rrdtool library version. > > >> 2 and 3 pping and servicemon failed to start, I tried a nav start pping >> servicemon a couple of time, also a start separate. >> but they whon't start. I changed DEBUG_LEVEL = 6 in nav.conf, but can't >> find any reason in the log files. >> > > The DEBUG_LEVEL in nav.conf only sets debug levels for the Java subsystems > (i.e. getDeviceData, eventEngine, getBoksMacs, networkDiscovery). The pping > and servicemon processes have their own debug level settings in their > respective config files. > > That said, pping and servicemon share much of the same code base, and they are > both known to have startup difficulties. We have yet to investigate what > causes the problem, but usually, when one experiences startup problems with > these two, it helps to retry it several times. > > > >From mels at mels-kooijman.nl Fri Sep 7 21:37:17 2007 From: mels at mels-kooijman.nl (Mels) Date: Fri Sep 7 20:38:51 2007 Subject: [Nav-users] Re: rrdtool and other problems on Suse 10.2 In-Reply-To: <46E19491.2050804(a)mels-kooijman.nl> References: <200709031022.52942.barbieri.marco(a)unimore.it> <46DFC491.4070702(a)uninett.no> <46E04EA1.3010906(a)mels-kooijman.nl> <46E13E6C.3020600(a)uninett.no> <46E19491.2050804(a)mels-kooijman.nl> Message-ID: <46E19A5D.7050305(a)mels-kooijman.nl> That problem solved, was ownership of a few files. Everything looks fine, start playing with nav. Mels Mels wrote: > Thanks for your reply. > > I removed the rrdtool, wich was from the suse media. > I compiled RRDtool 1.2.15 (the suse version was the same) and install it. > I recompiled psycopg-1.1.21, the cronjob produced a error. > > After this restartapache,tomcat,postgresql and nav > Even pping and servicemon started the first time. > > No errormail from the cronjobs, everythings looks fine, til i login in > nav: > > error > > File "/usr/local/nav/lib/python/nav/web/state.py", line 250, in __new__ > raise e > > Maybe i have to rebuild more modules? > > Mels > > Morten Brekkevold wrote: >> Mels wrote, On 09/06/2007 09:01 PM: >> >> >>> Hi all, >>> >> >> Hi Mels. >> >> >>> I started this week with a clean system, installing nav on Suse 10.2 >>> >>> There was a problem with connecting to a remote database, for the >>> time i >>> use the local database, witch is OK for nav. >>> >> >> Ok, if you need some tips on connecting to a remote database, please >> don't >> hesitate to ask :) >> >> >>> I get 3 problems: >>> 1 I installed py-rrdtool-1.0b1, with the installation off >>> py-rrdtool-0.2.1 I get a error: >>> building 'rrdtool._rrdtool' extension >>> gcc -pthread -fno-strict-aliasing -DNDEBUG -O2 -march=i586 -mtune=i686 >>> -fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -fPIC >>> -I/usr/local/include -I/usr/include/python2.5 -c >>> src/_rrdtoolmodule.c -o >>> build/temp.linux-i686-2.5/src/_rrdtoolmodule.o >>> src/_rrdtoolmodule.c: In function ?PyRRD_graph?: >>> src/_rrdtoolmodule.c:239: error: too few arguments to function >>> ?rrd_graph? >>> error: command 'gcc' failed with exit status 1 >>> >> >> I did not recognize the version numbers you mentioned, and checking >> up on >> things seems to reveal that there is some confusion about which >> python-rrdtool >> library to use. Apparently there are two very similar libraries in >> existance: >> One is distributed alongside rrdtool itself, and one is available at >> http://sourceforge.net/projects/py-rrdtool . >> >> The one currently in use by us on our dev server is the python-rrd >> library >> that comes with rrdtool itself; in Debian this package is called >> python-rrd. >> The package also follows the same version numbering scheme as >> rrdtool, which >> means that the latest version of it is >= 1.2. Unless Suse 10.2 >> prepackages >> this alongside its regular rrdtool package, you'll need to get hold >> of the >> rrdtool source code instead. This would also be the best way to make >> sure the >> python library version matches the rrdtool library version. >> >> >>> 2 and 3 pping and servicemon failed to start, I tried a nav start >>> pping >>> servicemon a couple of time, also a start separate. >>> but they whon't start. I changed DEBUG_LEVEL = 6 in nav.conf, but can't >>> find any reason in the log files. >>> >> >> The DEBUG_LEVEL in nav.conf only sets debug levels for the Java >> subsystems >> (i.e. getDeviceData, eventEngine, getBoksMacs, networkDiscovery). >> The pping >> and servicemon processes have their own debug level settings in their >> respective config files. >> >> That said, pping and servicemon share much of the same code base, and >> they are >> both known to have startup difficulties. We have yet to investigate >> what >> causes the problem, but usually, when one experiences startup >> problems with >> these two, it helps to retry it several times. >> >> >> > _______________________________________________ > nav-users mailing list > nav-users(a)itea.ntnu.no > http://mailman.itea.ntnu.no/mailman/listinfo/nav-users > >From ghitzacatalin at yahoo.com Sun Sep 9 14:33:57 2007 From: ghitzacatalin at yahoo.com (Gheorghe Limboi) Date: Sun Sep 9 22:34:06 2007 Subject: [Nav-users] IP address scope - graphical view / Traffic map load Message-ID: <362473.55698.qm(a)web50011.mail.re2.yahoo.com> Hi, regarding.. > 2. On traffic map -- there is no load on the link (Load: <0.01% No Data) > but when I follow the link (right clik - Load - this open the correct > cricket graph .. and thre is load on the link ) I fond this in tomcat server logs: SQLException for update statement: SELECT netboxid,descr,name,path,filename,key,value FROM rrd_file JOIN rrd_datasource USING(rrd_fi leid) WHERE subsystem='cricket' AND descr IN ('ifOutOctets','ifInOctets') AND ((key='gwport' AND value IN (33,24,10))) org.postgresql.util.PSQLException: ERROR: IN types character varying and integer cannot be matched This seems to be related to postgresql 8.2.x restriction ( I have postgresql-8.2.4 installed). Is there any work around or the only way is to downgrade to a lower version of postgresql tx, Gheorghe ----- Original Message ---- From: Morten Brekkevold <morten.brekkevold(a)uninett.no> To: Gheorghe Limboi <ghitzacatalin(a)yahoo.com> Cc: nav-users(a)itea.ntnu.no Sent: Friday, August 31, 2007 8:40:04 AM Subject: Re: [Nav-users] IP address scope - graphical view / Traffic map load Gheorghe Limboi wrote, On 24-08-2007 22:42: > Hi, Hi Gheorghe! > 1. Reports - > IP address scope - graphical view > > the olny result i get on the page is: 0326496128160192224 What exactly do you mean by "result"? Does the page consist of only this plaintext number and nothing else? > 2. On traffic map -- there is no load on the link (Load: <0.01% No Data) > but when I follow the link (right clik - Load - this open the correct > cricket graph .. and thre is load on the link ) > > > Any iddeas what I'm missing? You don't specify which version of NAV you are running, but I'm sure you're not using anything earlier than 3.1.1. We fixed a long standing problem with missing link load in the Traffic Map in that version, but you may be experiencing a different problem. The Traffic Map applet requests its data from the vPServer servlet running in the Tomcat container; the vPServer component calls the vprrd.py script to retrieve the actual RRD data, and something might have gone wrong during the calling of this script. I have to go now, but I will get back to you on this later, when I figure out whether there are any logs you can check to find the problem. -- mvh Morten Brekkevold UNINETT ____________________________________________________________________________________ Be a better Globetrotter. Get better travel answers from someone who knows. Yahoo! Answers - Check it out. http://answers.yahoo.com/dir/?link=list&sid=396545469 >From morten.brekkevold at uninett.no Tue Sep 11 14:41:33 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Tue Sep 11 13:41:37 2007 Subject: [Nav-users] IP address scope - graphical view / Traffic map load In-Reply-To: <362473.55698.qm(a)web50011.mail.re2.yahoo.com> References: <362473.55698.qm(a)web50011.mail.re2.yahoo.com> Message-ID: <46E67EED.1090604(a)uninett.no> Gheorghe Limboi wrote, On 09-09-2007 22:33: > I fond this in tomcat server logs: > > SQLException for update statement: SELECT netboxid,descr,name,path,filename,key,value FROM rrd_file JOIN rrd_datasource USING(rrd_fi > leid) WHERE subsystem='cricket' AND descr IN ('ifOutOctets','ifInOctets') AND ((key='gwport' AND value IN (33,24,10))) > org.postgresql.util.PSQLException: ERROR: IN types character varying and integer cannot be matched Ah, that explains why you're not getting any load data on your traffic map! > This seems to be related to postgresql 8.2.x restriction ( I have > postgresql-8.2.4 installed). Is there any work around or the only way is > to downgrade to a lower version of postgresql I don't see any workaround other than downgrading PostgreSQL :-( I've just looked at the code that creates the offending SQL statement - it looks hairy, but I'll try to fix it. I can post the patch here when it's done, if you like. -- mvh Morten Brekkevold UNINETT >From morten.brekkevold at uninett.no Tue Sep 11 14:45:30 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Tue Sep 11 13:47:41 2007 Subject: [Nav-users] Re: rrdtool and other problems on Suse 10.2 In-Reply-To: <46E19A5D.7050305(a)mels-kooijman.nl> References: <200709031022.52942.barbieri.marco(a)unimore.it> <46DFC491.4070702(a)uninett.no> <46E04EA1.3010906(a)mels-kooijman.nl> <46E13E6C.3020600(a)uninett.no> <46E19491.2050804(a)mels-kooijman.nl> <46E19A5D.7050305(a)mels-kooijman.nl> Message-ID: <46E67FDA.3030101(a)uninett.no> Mels wrote, On 07-09-2007 20:37: > That problem solved, was ownership of a few files. > > Everything looks fine, start playing with nav. Great! Just give a shout if there are any more problems :) -- mvh Morten Brekkevold UNINETT >From marco.barbieri at unimore.it Mon Sep 17 15:41:30 2007 From: marco.barbieri at unimore.it (Marco Barbieri) Date: Mon Sep 17 14:41:39 2007 Subject: [Nav-users] NAV & VMware Message-ID: <200709171441.30621.marco.barbieri(a)unimore.it> Hi all, since i'm running my copies of NAV inside VMware machines i'm here to ask if could be of any help to the developing staff do give them a copy of an abstract machine. The main advantage that i could see would be skipping through the "Installing on Centos 4.4" faq, reconfiguration would be limited at network parameters, the generation of SSL certificate and... and what? Let me know if this sounds interesting. Best reguards, Marco Barbieri. -- Marco Barbieri - CeSIA - 059-2055218

1 0

Centos 5 and NAV 3.2.2
by barbieri.marco＠unimore.it 03 Sep '07

03 Sep '07

Hi, first of all i'm trying to run NAV 3.2.2 on a Centos 5 and i know it's not supported, but with really liitle tweaking the system seem to be up and running, despite the problem i've always had with my HP9304m that i've already posted about (answer pending...). I get some errors from: Cron <navcron@nav> $BINDIR/getBoksMacs.sh $NUM_THREADS the body of the message is at the end of this msg, there's something i can do to fix it or the set of libs that comes with Centos5 is too much different to let NAV work as it is? Thanks for reading and my best regards, Marco barbieri. ---Body--- org.postgresql.util.PSQLException: This statement has been closed. at org.postgresql.jdbc2.AbstractJdbc2Statement.checkClosed(postgresql-jdbc-8.1.407.jar.so) at org.postgresql.jdbc2.AbstractJdbc2Statement.getUpdateCount(postgresql-jdbc-8.1.407.jar.so) at org.postgresql.jdbc2.AbstractJdbc2Statement.executeUpdate(postgresql-jdbc-8.1.407.jar.so) at no.ntnu.nav.Database.Database.update(Database.java:1004) at no.ntnu.nav.Database.Database.insert(Database.java:928) at QueryBoks.runCamQueue(QueryBoks.java:1659) at QueryBoks.processMacEntry(QueryBoks.java:977) at QueryBoks.run(QueryBoks.java:266) SQLException for update statement: INSERT INTO cam (netboxid,sysname,ifindex,module,port,mac,start_time) VALUES ('6','rett-mo-hp.unimo.it','128','0','28','0011d80d2cde',NOW()) org.postgresql.util.PSQLException: This statement has been closed. at org.postgresql.jdbc2.AbstractJdbc2Statement.checkClosed(postgresql-jdbc-8.1.407.jar.so) at org.postgresql.jdbc2.AbstractJdbc2Statement.getUpdateCount(postgresql-jdbc-8.1.407.jar.so) at org.postgresql.jdbc2.AbstractJdbc2Statement.executeUpdate(postgresql-jdbc-8.1.407.jar.so) at no.ntnu.nav.Database.Database.update(Database.java:1004) at no.ntnu.nav.Database.Database.insert(Database.java:928) at QueryBoks.runCamQueue(QueryBoks.java:1659) at QueryBoks.processMacEntry(QueryBoks.java:977) at QueryBoks.run(QueryBoks.java:266) ---Body--- -- Marco Barbieri - CeSIA - 059-2055218 >From morten.brekkevold at uninett.no Thu Sep 6 11:35:14 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Thu Sep 6 10:35:38 2007 Subject: [Nav-users] Problem with NAV 3.2.2 and Hp 9304M - WRONG DATA - SORRY In-Reply-To: <200708310840.46033.barbieri.marco(a)unimore.it> References: <200708310840.46033.barbieri.marco(a)unimore.it> Message-ID: <46DFBBC2.2050000(a)uninett.no> Marco Barbieri wrote, On 31-08-2007 08:40: > Hi again, > i'm sorry i've posted the wrong data, in the previous post you can find a > HP5412zl that NAV detects well. > > Here there's the data about HP9304M (it's marked HP but it's a Foundry > machine) that has the problem, it's a chassis with 4 slots, 3 of them are > loaded with 8 gbic sockets boards. Given the OID compatibility report you sent, NAV should be able to detect the ports on this routing switch; the ports are detected using standard mibs. Yet you say that _no_ ports are detected? If that is so, I would like to see what getDeviceData.log says when it collects from this device. When it comes to modules, I must admit that NAV's support for HP equipment is lacking in several respects. The HP support was written and tested using the equipment that was available at the Norwegian University of Science and Technology at the time, and these were all HP SwitchStacks. Consequently, NAV only recognizes virtual stacking in HP equipment, and cannot at all recognize modules in your HP chassis. It is in our plans to implement more generic module support through interpreting the ENTITY-MIB. I'm guessing your HP9304M presents a fairly correct view of its physical contents through its entPhysicalTable? Since I have no access to a 9304M myself, but am planning to code the ENTITY-MIB support, I would greatly appreciate it if you could provide me with the output of this table from your device; running this command on your NAV server should do it: snmptable -Ci -Cb -v2c -c <community> <ip> entPhysicalTable -- mvh Morten Brekkevold UNINETT >From morten.brekkevold at uninett.no Thu Sep 6 12:12:49 2007 From: morten.brekkevold at uninett.no (Morten Brekkevold) Date: Thu Sep 6 11:13:05 2007 Subject: [Nav-users] Centos 5 and NAV 3.2.2 In-Reply-To: <200709031022.52942.barbieri.marco(a)unimore.it> References: <200709031022.52942.barbieri.marco(a)unimore.it> Message-ID: <46DFC491.4070702(a)uninett.no> Marco Barbieri wrote, On 03-09-2007 10:22: > first of all i'm trying to run NAV 3.2.2 on a Centos 5 and i know it's not > supported, That depends on what you mean by "supported" :) We do not guaruantee support to anyone, but you should be able to have NAV up and run on any Linux distro. The most common obstacle seems to be when some software that NAV depends on comes out with a new version that breaks some backwards compatibility. > I get some errors from: > > Cron <navcron@nav> $BINDIR/getBoksMacs.sh $NUM_THREADS > > the body of the message is at the end of this msg, there's something i can do > to fix it or the set of libs that comes with Centos5 is too much different to > let NAV work as it is? > > Thanks for reading and my best regards, Marco barbieri. > > ---Body--- > org.postgresql.util.PSQLException: This statement has been closed. > at > org.postgresql.jdbc2.AbstractJdbc2Statement.checkClosed(postgresql-jdbc-8.1.407.jar.so) [...] I think I've seen this before, and it happened when we upgraded our PostgreSQL JDBC driver to version 8.1. I'm not sure what changed, since the API should remain the same - only the driver should work better against PostgreSQL 8.1. I assume that your CentOS installation runs the PostgreSQL 8.1 server as well? Our temporary workaround was to downgrade the JDBC driver to the 7.4 version. You could try and see if that works. If CentOS 5 doesn't provide that version of the driver, you can get it from http://jdbc.postgresql.org/ -- mvh Morten Brekkevold UNINETT >From mels at mels-kooijman.nl Thu Sep 6 22:01:53 2007 From: mels at mels-kooijman.nl (Mels) Date: Thu Sep 6 21:02:09 2007 Subject: [Nav-users] Centos 5 and NAV 3.2.2 In-Reply-To: <46DFC491.4070702(a)uninett.no> References: <200709031022.52942.barbieri.marco(a)unimore.it> <46DFC491.4070702(a)uninett.no> Message-ID: <46E04EA1.3010906(a)mels-kooijman.nl> Hi all, I started this week with a clean system, installing nav on Suse 10.2 There was a problem with connecting to a remote database, for the time i use the local database, witch is OK for nav. I get 3 problems: 1 I installed py-rrdtool-1.0b1, with the installation off py-rrdtool-0.2.1 I get a error: building 'rrdtool._rrdtool' extension gcc -pthread -fno-strict-aliasing -DNDEBUG -O2 -march=i586 -mtune=i686 -fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -fPIC -I/usr/local/include -I/usr/include/python2.5 -c src/_rrdtoolmodule.c -o build/temp.linux-i686-2.5/src/_rrdtoolmodule.o src/_rrdtoolmodule.c: In function ?PyRRD_graph?: src/_rrdtoolmodule.c:239: error: too few arguments to function ?rrd_graph? error: command 'gcc' failed with exit status 1 from the cron I get a mail with the next text: File "/usr/local/nav/bin/thresholdMon.py", line 46, in <module> from nav.rrd import presenter File "/usr/local/nav/lib/python/nav/rrd/presenter.py", line 54, in <module> import rrdtool ImportError: dynamic module does not define init function (initrrdtool) 2 and 3 pping and servicemon failed to start, I tried a nav start pping servicemon a couple of time, also a start separate. but they whon't start. I changed DEBUG_LEVEL = 6 in nav.conf, but can't find any reason in the log files. Mels

1 0

psycopg2._psycopg.connection
by mels＠mels-kooijman.nl 06 Aug '07

06 Aug '07

Hi, I'm new to nav and have installed nav from source on OpenSuse 10.2 As a guide i keep de suse installation guide. When i want to connect with the browser i get an error, in the apachelog is the same error: [Sun Aug 05 19:57:37 2007] [error] [client 192.168.2.3] PythonCleanupHandler nav.web: File "/usr/local/nav/lib/python/nav/db/__init__.py", line 109, in getConnection\n connection.autocommit(0) [Sun Aug 05 19:57:37 2007] [error] [client 192.168.2.3] PythonCleanupHandler nav.web: AttributeError: 'psycopg2._psycopg.connection' object has no attribute 'autocommit' Anybody has a idea whats wrong? Greetings, Mels

1 1

Mailing list problems
by morten.brekkevold＠uninett.no 17 Jul '07

17 Jul '07

We still seem to be having problems with our mailing list setup. nav-commit and nav-dev are confirmed to be working fine, but nav-users apparently doesn't. Some subscribers seem to be getting the mails sent to nav-dev(a)itea.ntnu.no, while others are not (myself included). However, none of these mails are stored in the mailing list archives. I've asked the guys at itea.ntnu.no to take a look at this problem, but apparently all their postmasters have taken their summer vacations simultaineously (!), so I cannot really say when this problem will be fixed. I would have moved the mailing lists to a different server to ensure continued service, were it not for the fact that I'll be taking 4 weeks of vacation starting today. Therefore I've asked John-Magne, the second list administrator, to try to influence things at itea.ntnu.no while I'm gone. Some of you might get this e-mail, and some of you might not, I don't know. If some of you who receive this e-mail can respond to it, maybe we can gather enough clues as to what is wrong that the postmasters (or acting postmasters or whatever) at itea.ntnu.no can fix the problem. If not, the list may be defunct for the next 4 weeks :-( -- mvh Morten Brekkevold UNINETT

2 1