From heas at shrubbery.net Tue Feb 2 01:49:14 2016 From: heas at shrubbery.net ('heasley') Date: Tue, 2 Feb 2016 01:49:14 +0000 Subject: [rancid] rancid : how to filter clogin -x or -c outcome In-Reply-To: <003901d07d9e$57e38e20$07aaaa60$@laposte.net> References: <000501d07a73$6a8fefb0$3fafcf10$@laposte.net> <20150420161254.GS88638@shrubbery.net> <003901d07d9e$57e38e20$07aaaa60$@laposte.net> Message-ID: <20160202014914.GA47590@shrubbery.net> Thu, Apr 23, 2015 at 10:20:21AM +0200, alligator94: > Thanks for the info. > I would need some help as I am unable to make par(1) work. > My tests run on a test system. > In the same directory as clogin, I have created a file named "infile" which > contains the command to run. It contains 2 lines > # ( to indicate that the next one is a command) > clogin -c "show users" hostname > > Then I run ./par infile. Nothing happens but 3 empty log files are > created(par.log....) > I have tried different combinations without success. I had a bug in quote handling. Please try ftp://ftp.shrubbery.net/pub/rancid/alpa/rancid-3.2.99.tar.gz which will likely be rancid 3.3 very soon. [sorry that it took so long to look into this problem.] > I have not found any sample except the one coming with the man page which > doesn't help me. > > Regards, > Gilles > > -----Original Message----- > From: heasley [mailto:heas at shrubbery.net] > Sent: lundi 20 avril 2015 18:13 > To: alligator94 > Cc: rancid-discuss at shrubbery.net > Subject: Re: [rancid] rancid : how to filter clogin -x or -c outcome > > Sun, Apr 19, 2015 at 09:35:30AM +0200, alligator94: > > Hi, > > > > > > > > I very often use clogin -x or clogin -c to perform mass configuration > > updates. My concern is that clogin doesn't analyze the outcome of the > > command passed with -x or -d, so there is no easy way to know if the > > command has been successfully accepted by the device . For instance, > > some old cisco ios don't accept some commands. > > > > I have tried to use the clogin debug (-d) option, but it generates a > > huge amount of lines. I have also tested to open a log file (I am > > using teraterm) before running clogin, then to do a search but, it is > > not really easy because each time that you find an error condition, > > you need to look backward in the file to know which host the error message > reports to. > > Would par(1) help? > > > I have also looked at the -s option but it looks complex to me. > > > > > > > > What would be nice to have is a way to pass a file to clogin as a > > parameter which would contain a list of keyword that you want to be > > extracted from the outcome. > > > > I am thinking to some words like : > > > > "Invalid" > > > > "Error" > > > > "timeout" > > extracted from the rest of the o/p in what manner? > > > If some of them are trapped during the clogin run, then they could be > > stored in a file $hostname_date_time.err.log > > I think par(1) would indeed help you here. you could egrep the stuff you > want to extract (or awk, etc). > > > > > > > Maybe there is an easy way to do it but I am not skilled with expect. > > > > Any help or comment is welcomed. > > > > > > > > Thanks > > > > Regards, > > > > Gilles > > > > > _______________________________________________ > > Rancid-discuss mailing list > > Rancid-discuss at shrubbery.net > > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From nachofw at vera.com.uy Tue Feb 2 14:42:46 2016 From: nachofw at vera.com.uy (Ignacio Freyre) Date: Tue, 2 Feb 2016 11:42:46 -0300 (UYT) Subject: [rancid] cisco asa "terminal length" In-Reply-To: <308014776.70733143.1454423992914.JavaMail.zimbra@vera.com.uy> Message-ID: <1711646167.70735957.1454424166849.JavaMail.zimbra@vera.com.uy> hi, i was looking at the clogin debug for an asa, i see that "terminal length 0" and "terminal width 132" commands fail. cisco asa uses "terminal pager 0". How can i correct this so my asa devices use the command "terminal pager 0". regards asa# terminal length 0 ^ ERROR: % Invalid input detected at '^' marker. asa# terminal width 132 ^ ERROR: % Invalid input detected at '^' marker. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nachofw at vera.com.uy Tue Feb 2 15:58:02 2016 From: nachofw at vera.com.uy (Ignacio Freyre) Date: Tue, 2 Feb 2016 12:58:02 -0300 (UYT) Subject: [rancid] html colorized mail diffs In-Reply-To: <568557908.70798946.1454428484664.JavaMail.zimbra@vera.com.uy> Message-ID: <131425330.70801722.1454428682159.JavaMail.zimbra@vera.com.uy> thanks heasley, ultimately i made it work just changing the rancid config file. the python script is much more easy to read, i went with that one. 1) chmod +x /var/lib/rancid/diff2html.py 2) config change -> /etc/rancid/rancid.conf MAILHEADERS="Content-Type: text/html; charset=ISO-8859-1"; export MAILHEADERS DIFFSCRIPT="/var/lib/rancid/diff2html.py"; export DIFFSCRIPT scripts (both of them work): http://stackoverflow.com/questions/3827862/convert-raw-diff-file-to-colorized-html-output http://www.linuxjournal.com/content/convert-diff-output-colorized-html -------------- next part -------------- An HTML attachment was scrubbed... URL: From nachofw at vera.com.uy Tue Feb 2 17:36:08 2016 From: nachofw at vera.com.uy (Ignacio Freyre) Date: Tue, 2 Feb 2016 14:36:08 -0300 (UYT) Subject: [rancid] cisco asa "terminal length" In-Reply-To: <1711646167.70735957.1454424166849.JavaMail.zimbra@vera.com.uy> References: <1711646167.70735957.1454424166849.JavaMail.zimbra@vera.com.uy> Message-ID: <1755377146.70881082.1454434568412.JavaMail.zimbra@vera.com.uy> i have created a copy of clogin, asaclogin. For now its working. Could this be corrected in upcoming releases of rancid? --- clogin 2016-01-11 15:02:44.318068528 -0300 +++ asaclogin 2016-02-02 13:41:45.156787795 -0300 @@ -999,11 +999,9 @@ send "set width 132\r" expect -re $prompt {} send "set logging session disable\r" } else { - send "terminal length 0\r" - expect -re $prompt {} - send "terminal width 132\r" + send "terminal pager 0\r" } expect -re $prompt {} } else { send "disable clipaging\r" -------------- next part -------------- An HTML attachment was scrubbed... URL: From nachofw at vera.com.uy Wed Feb 3 16:36:23 2016 From: nachofw at vera.com.uy (Ignacio Freyre) Date: Wed, 3 Feb 2016 13:36:23 -0300 (UYT) Subject: [rancid] cvs time in UTC In-Reply-To: <2014073421.71605556.1454516667252.JavaMail.zimbra@vera.com.uy> Message-ID: <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> Hi, i noticed cvs stores time info in UTC time. Is there a way to change to the system time which uses gmt? In this example the actual system time was 9:41:39, but cvs stored the time as 12:41:39, i got the email with the diff but the time is wrong. cvs diff -u -4 -r1.15 configs/test-router Index: configs/test-router =================================================================== RCS file: /var/lib/rancid/CVS/prueba/configs/test-router,v retrieving revision 1.15 retrieving revision 1.16 diff -u -4 -r1.15 -r1.16 --- configs/test-router 3 Feb 2016 12:41:39 -0000 1.15 +++ configs/test-router 3 Feb 2016 14:51:59 -0000 1.16 under CVS root: 1.16 date 2016.02.03.14.51.59; author rancid; state Exp; branches; next 1.15; commitid 10056B2140F607F79E7; -------------- next part -------------- An HTML attachment was scrubbed... URL: From nachofw at vera.com.uy Wed Feb 3 17:28:24 2016 From: nachofw at vera.com.uy (Ignacio Freyre) Date: Wed, 3 Feb 2016 14:28:24 -0300 (UYT) Subject: [rancid] cvs time in UTC In-Reply-To: References: <2014073421.71605556.1454516667252.JavaMail.zimbra@vera.com.uy> <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> Message-ID: <1949840288.71652792.1454520504569.JavaMail.zimbra@vera.com.uy> the system time is fine, i'm using UTC-3 or GMT-3. What i meant is that cvs is using UTC-0, but the system time is UTC-3. ----- Mensaje original ----- De: "Aaron Dudek" Para: "Ignacio Freyre" CC: rancid-discuss at shrubbery.net Enviados: Mi?rcoles, 3 de Febrero 2016 14:17:24 Asunto: Re: [rancid] cvs time in UTC Why do you have a three hour difference? I think that is the bigger question, On Wednesday, February 3, 2016, Ignacio Freyre < nachofw at vera.com.uy > wrote: Hi, i noticed cvs stores time info in UTC time. Is there a way to change to the system time which uses gmt? In this example the actual system time was 9:41:39, but cvs stored the time as 12:41:39, i got the email with the diff but the time is wrong. cvs diff -u -4 -r1.15 configs/test-router Index: configs/test-router =================================================================== RCS file: /var/lib/rancid/CVS/prueba/configs/test-router,v retrieving revision 1.15 retrieving revision 1.16 diff -u -4 -r1.15 -r1.16 --- configs/test-router 3 Feb 2016 12:41:39 -0000 1.15 +++ configs/test-router 3 Feb 2016 14:51:59 -0000 1.16 under CVS root: 1.16 date 2016.02.03.14.51.59; author rancid; state Exp; branches; next 1.15; commitid 10056B2140F607F79E7; -------------- next part -------------- An HTML attachment was scrubbed... URL: From heas at shrubbery.net Wed Feb 3 17:29:26 2016 From: heas at shrubbery.net (heasley) Date: Wed, 3 Feb 2016 17:29:26 +0000 Subject: [rancid] cvs time in UTC In-Reply-To: <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> References: <2014073421.71605556.1454516667252.JavaMail.zimbra@vera.com.uy> <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> Message-ID: <20160203172926.GE15940@shrubbery.net> Wed, Feb 03, 2016 at 01:36:23PM -0300, Ignacio Freyre: > Hi, i noticed cvs stores time info in UTC time. Is there a way to change to the system time which uses gmt? > In this example the actual system time was 9:41:39, but cvs stored the time as 12:41:39, i got the email with the diff but the time is wrong. > It uses the system timezone by default. you can change the global system timezone (do not ask this list how to do that) or in your rancid.conf and/or shell dot files often by setting the TZ environment variable (again, do not ask here about how to do that - google or ask a list associated with your O/S). From adudek16 at gmail.com Wed Feb 3 17:17:24 2016 From: adudek16 at gmail.com (Aaron Dudek) Date: Wed, 3 Feb 2016 12:17:24 -0500 Subject: [rancid] cvs time in UTC In-Reply-To: <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> References: <2014073421.71605556.1454516667252.JavaMail.zimbra@vera.com.uy> <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> Message-ID: Why do you have a three hour difference? I think that is the bigger question, On Wednesday, February 3, 2016, Ignacio Freyre wrote: > Hi, i noticed cvs stores time info in UTC time. Is there a way to change > to the system time which uses gmt? > In this example the actual system time was 9:41:39, but cvs stored the > time as 12:41:39, i got the email with the diff but the time is wrong. > > > cvs diff -u -4 -r1.15 configs/test-router > > Index: configs/test-router > =================================================================== > RCS file: /var/lib/rancid/CVS/prueba/configs/test-router,v > retrieving revision 1.15 > retrieving revision 1.16 > diff -u -4 -r1.15 -r1.16 > --- configs/test-router 3 Feb 2016 12:41:39 -0000 1.15 > +++ configs/test-router 3 Feb 2016 14:51:59 -0000 1.16 > > > > under CVS root: > > 1.16 > date 2016.02.03.14.51.59; author rancid; state Exp; > branches; > next 1.15; > commitid 10056B2140F607F79E7; > -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.ahrens at citrix.com Wed Feb 3 18:37:45 2016 From: david.ahrens at citrix.com (David Ahrens) Date: Wed, 3 Feb 2016 18:37:45 +0000 Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> References: <642FAF39868FD54A909CBFC1DD8F61A70D6D6309@SJCPEX01CL02.citrite.net> <20160127072712.GC33037@shrubbery.net> <56A8967A.9000007@thingy.com> <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> Message-ID: <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> Hi Elliot, I have some old Brocade 200 SAN switches running version 5.2.2 of Fabric OS. The brocaderancid script seems to run properly but for some reason doesn't recognize/capture the out of the commands. The attached log file shows the output of the switch commands, any ideas? Thanks, David -----Original Message----- From: Rancid-discuss [mailto:rancid-discuss-bounces at shrubbery.net] On Behalf Of Elliot Wilen Sent: Friday, January 29, 2016 11:00 AM To: rancid-discuss at shrubbery.net Subject: Re: [rancid] rancid support for Brocade Fabric OS Comparing Howard's version of the script to mine, I see a few differences, one of which may be a modification I made. Instead of using configShow directly, my script uses configUpload -all and saves it to a local file on the switch, followed by configShow -l to display that file. I think this yields a more complete configuration similar to what one would use to back up the switch via ftp. His version also has a couple additions relative to whatever the common source was. _______________________________________________ Rancid-discuss mailing list Rancid-discuss at shrubbery.net http://www.shrubbery.net/mailman/listinfo/rancid-discuss -------------- next part -------------- A non-text attachment was scrubbed... Name: brocadelogin Type: application/octet-stream Size: 26636 bytes Desc: brocadelogin URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: brocaderancid Type: application/octet-stream Size: 16294 bytes Desc: brocaderancid URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: EI.20160203.101713 Type: application/octet-stream Size: 25292 bytes Desc: EI.20160203.101713 URL: From nachofw at vera.com.uy Wed Feb 3 19:09:33 2016 From: nachofw at vera.com.uy (Ignacio Freyre) Date: Wed, 3 Feb 2016 16:09:33 -0300 (UYT) Subject: [rancid] cvs time in UTC In-Reply-To: <20160203172926.GE15940@shrubbery.net> References: <2014073421.71605556.1454516667252.JavaMail.zimbra@vera.com.uy> <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> <20160203172926.GE15940@shrubbery.net> Message-ID: <1689053912.71736892.1454526573815.JavaMail.zimbra@vera.com.uy> Hi heasley, the timezone is configured properly. Still, rancid logs show the correct time, but cvs doesn't. I have configured the TZ environment variable but that didn't change anything. Rancid log from running 'rancid-run': starting: Wed Feb 3 16:05:55 UYT 2016 CVS root: 1.22 date 2016.02.03.19.06.01; author rancid; state Exp; branches; next 1.21; commitid 10056B24F996FB6D309; rancid at bckp-server:~$ timedatectl Local time: Wed 2016-02-03 15:34:16 UYT Universal time: Wed 2016-02-03 18:34:16 UTC Timezone: America/Montevideo (UYT, -0300) NTP enabled: yes NTP synchronized: yes RTC in local TZ: no DST active: n/a ----- Mensaje original ----- De: "heasley" Para: "Ignacio Freyre" CC: rancid-discuss at shrubbery.net Enviados: Mi?rcoles, 3 de Febrero 2016 14:29:26 Asunto: Re: [rancid] cvs time in UTC Wed, Feb 03, 2016 at 01:36:23PM -0300, Ignacio Freyre: > Hi, i noticed cvs stores time info in UTC time. Is there a way to change to the system time which uses gmt? > In this example the actual system time was 9:41:39, but cvs stored the time as 12:41:39, i got the email with the diff but the time is wrong. > It uses the system timezone by default. you can change the global system timezone (do not ask this list how to do that) or in your rancid.conf and/or shell dot files often by setting the TZ environment variable (again, do not ask here about how to do that - google or ask a list associated with your O/S). From heas at shrubbery.net Wed Feb 3 21:21:23 2016 From: heas at shrubbery.net (heasley) Date: Wed, 3 Feb 2016 21:21:23 +0000 Subject: [rancid] cvs time in UTC In-Reply-To: <1689053912.71736892.1454526573815.JavaMail.zimbra@vera.com.uy> References: <2014073421.71605556.1454516667252.JavaMail.zimbra@vera.com.uy> <530432128.71614102.1454517383109.JavaMail.zimbra@vera.com.uy> <20160203172926.GE15940@shrubbery.net> <1689053912.71736892.1454526573815.JavaMail.zimbra@vera.com.uy> Message-ID: <20160203212123.GJ15940@shrubbery.net> Wed, Feb 03, 2016 at 04:09:33PM -0300, Ignacio Freyre: > Hi heasley, the timezone is configured properly. > Still, rancid logs show the correct time, but cvs doesn't. > I have configured the TZ environment variable but that didn't change anything. > > > Rancid log from running 'rancid-run': starting: Wed Feb 3 16:05:55 UYT 2016 > CVS root: > 1.22 > date 2016.02.03.19.06.01; author rancid; state Exp; > branches; > next 1.21; > commitid 10056B24F996FB6D309; i've confused this with svn. cvs does not do tz translation except for the history command, afaict from the cvs manpage. > rancid at bckp-server:~$ timedatectl > Local time: Wed 2016-02-03 15:34:16 UYT > Universal time: Wed 2016-02-03 18:34:16 UTC > Timezone: America/Montevideo (UYT, -0300) > NTP enabled: yes > NTP synchronized: yes > RTC in local TZ: no > DST active: n/a > > > ----- Mensaje original ----- > De: "heasley" > Para: "Ignacio Freyre" > CC: rancid-discuss at shrubbery.net > Enviados: Mi?rcoles, 3 de Febrero 2016 14:29:26 > Asunto: Re: [rancid] cvs time in UTC > > Wed, Feb 03, 2016 at 01:36:23PM -0300, Ignacio Freyre: > > Hi, i noticed cvs stores time info in UTC time. Is there a way to change to the system time which uses gmt? > > In this example the actual system time was 9:41:39, but cvs stored the time as 12:41:39, i got the email with the diff but the time is wrong. > > > > It uses the system timezone by default. you can change the global system > timezone (do not ask this list how to do that) or in your rancid.conf > and/or shell dot files often by setting the TZ environment variable (again, > do not ask here about how to do that - google or ask a list associated with > your O/S). From Dan.Weintraub at dealer.com Fri Feb 5 22:28:31 2016 From: Dan.Weintraub at dealer.com (Dan Weintraub) Date: Fri, 5 Feb 2016 22:28:31 +0000 Subject: [rancid] rancid git setup question Message-ID: <98193ABC-A665-4E5D-B7AB-BCEBA29136EA@dealer.com> Hi, I?m a first time user having a question about making rancid work with git. I?m on CentOS 7 with rancid 3.3.0, perl v5.16.3, and git 1.8.3.1. I read that "no attempt will be made to create the repository when running rancid-cvs? so I cloned my empty git repo into the rancid $BASEDIR and pointed at it in the rancid.conf: CVSROOT=$BASEDIR/rancid-backups/; export CVSROOT I then ran rancid-cvs, but I don?t understand the behavior. I would expect it to prepare the CVSROOT, but it made a new folder and repo named after my sole group. It then made a new folder in my intended git repo ($CVSROOT), with my group name, but it looks like a .git directory: [rancid at rancid rancid-backups]$ ls -al total 20 ... drwxr-x--- 7 rancid rancid 4096 Feb 5 22:01 myGroup drwxr-xr-x 8 rancid rancid 4096 Feb 5 22:01 .git -rw-r--r-- 1 rancid rancid 60 Feb 5 22:01 README.md [rancid at rancid rancid-backups]$ ls -al myGroup/ total 40 ... drwxr-x--- 2 rancid rancid 4096 Feb 5 22:01 branches -rw-r----- 1 rancid rancid 66 Feb 5 22:01 config -rw-r----- 1 rancid rancid 73 Feb 5 22:01 description -rw-r----- 1 rancid rancid 23 Feb 5 22:01 HEAD drwxr-x--- 2 rancid rancid 4096 Feb 5 22:01 hooks drwxr-x--- 2 rancid rancid 4096 Feb 5 22:01 info drwxr-x--- 11 rancid rancid 4096 Feb 5 22:01 objects drwxr-x--- 4 rancid rancid 4096 Feb 5 22:01 refs When I run rancid-run, it puts the config files in the myGroup directory of the BASEDIR, but it also seems to touch some objects in the myGroup .git directory. I then thought that I would only use the repo generated my rancid-cvs, so I put the CVSROOT back to default and tried again. It still makes a $CVSROOT/myGroup directory that doesn?t seem to make sense. I would appreciate it if someone could explain what?s going on or recommend some good documentation on getting rancid and git set up. Thanks, Dan Dan Weintraub | Systems Engineer V: 877.327.8422 x mailto:Dan.Weintraub at dealer.com | www.dealer.com From heas at shrubbery.net Sat Feb 6 01:09:42 2016 From: heas at shrubbery.net (heasley) Date: Sat, 6 Feb 2016 01:09:42 +0000 Subject: [rancid] rancid git setup question In-Reply-To: <98193ABC-A665-4E5D-B7AB-BCEBA29136EA@dealer.com> References: <98193ABC-A665-4E5D-B7AB-BCEBA29136EA@dealer.com> Message-ID: <20160206010942.GA73515@shrubbery.net> Fri, Feb 05, 2016 at 10:28:31PM +0000, Dan Weintraub: > Hi, > > I?m a first time user having a question about making rancid work with git. I?m on CentOS 7 with rancid 3.3.0, perl v5.16.3, and git 1.8.3.1. > > I read that "no attempt will be made to create the repository when running rancid-cvs? so I cloned my empty git repo into the rancid $BASEDIR and pointed at it in the rancid.conf: I need to make that clearer; no attempt will be made to create repos if CVSROOT looks like a URL. the script does not handle remote repos. ie: if you want the origin (if i have the parlance correct) to be on a remote host, you must create it yourself. My suggestion would be that you let rancid-cvs create the repo locally, then move that to your remote host and update group/.git/config to reflect that. CVSROOT does not translate well to git; for rancid's purpose, it puts all of it origins - one per group - under CVSROOT. so CVSROOT/myGroup, CVSROOT/youGroup, CVSROOT/groupies etc. And, the working directory is BASEDIR/myGroup. lmk if thats muddy ... or if i have a bug. > CVSROOT=$BASEDIR/rancid-backups/; export CVSROOT > > I then ran rancid-cvs, but I don?t understand the behavior. I would expect it to prepare the CVSROOT, but it made a new folder and repo named after my sole group. It then made a new folder in my intended git repo ($CVSROOT), with my group name, but it looks like a .git directory: > > [rancid at rancid rancid-backups]$ ls -al > total 20 > ... > drwxr-x--- 7 rancid rancid 4096 Feb 5 22:01 myGroup > drwxr-xr-x 8 rancid rancid 4096 Feb 5 22:01 .git > -rw-r--r-- 1 rancid rancid 60 Feb 5 22:01 README.md > [rancid at rancid rancid-backups]$ ls -al myGroup/ > total 40 > ... > drwxr-x--- 2 rancid rancid 4096 Feb 5 22:01 branches > -rw-r----- 1 rancid rancid 66 Feb 5 22:01 config > -rw-r----- 1 rancid rancid 73 Feb 5 22:01 description > -rw-r----- 1 rancid rancid 23 Feb 5 22:01 HEAD > drwxr-x--- 2 rancid rancid 4096 Feb 5 22:01 hooks > drwxr-x--- 2 rancid rancid 4096 Feb 5 22:01 info > drwxr-x--- 11 rancid rancid 4096 Feb 5 22:01 objects > drwxr-x--- 4 rancid rancid 4096 Feb 5 22:01 refs > > > When I run rancid-run, it puts the config files in the myGroup directory of the BASEDIR, but it also seems to touch some objects in the myGroup .git directory. > > I then thought that I would only use the repo generated my rancid-cvs, so I put the CVSROOT back to default and tried again. It still makes a $CVSROOT/myGroup directory that doesn?t seem to make sense. I would appreciate it if someone could explain what?s going on or recommend some good documentation on getting rancid and git set up. > > Thanks, > Dan > > > > > > Dan Weintraub | Systems Engineer > V: 877.327.8422 x > mailto:Dan.Weintraub at dealer.com | www.dealer.com > > > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From matthew at walster.org Sat Feb 6 02:20:53 2016 From: matthew at walster.org (Matthew Walster) Date: Fri, 5 Feb 2016 21:20:53 -0500 Subject: [rancid] rancid source In-Reply-To: <56A8F0B1.4070502@nexellent.ch> References: <56A8F0B1.4070502@nexellent.ch> Message-ID: On 27 January 2016 at 11:30, Roman Hochuli wrote: > Hello Jason > > > Is the RANCiD source being maintained on githug or are there any plans > > to migrate to github? It would certainly make it easier to submit > > patches and discuss issues. > > Have a look here: https://github.com/dotwaffle/rancid-git ?So, I've been the maintainer for that fork for a while now. A combination of a gross lack of free time, git being implemented (in a different way) upstream, and it being well over a year out of date has meant that this repo is basically dead now. I merge stuff when asked, but I ought to stop that. Within the next few weeks, I'll probably either delete the dotwaffle/rancid-git repo or add a commit that deletes everything and leaves a README of where to get the latest code -- if someone else wants to take it over, feel free. I imagine Mr. Heasley would rather you use his releases, however ;) M ? -------------- next part -------------- An HTML attachment was scrubbed... URL: From dale.shaw+rancid-discuss at gmail.com Sat Feb 6 02:43:39 2016 From: dale.shaw+rancid-discuss at gmail.com (Dale Shaw) Date: Sat, 6 Feb 2016 13:43:39 +1100 Subject: [rancid] rancid source In-Reply-To: References: <56A8F0B1.4070502@nexellent.ch> Message-ID: Hi all, On Sat, Feb 6, 2016 at 1:20 PM, Matthew Walster wrote: > [...] > Within the next few weeks, I'll probably either delete the dotwaffle/rancid-git repo or add a commit that deletes everything and leaves a README of where to get the latest code -- if someone else wants to take it over, feel free. I imagine Mr. Heasley would rather you use his releases, however ;) Ahuh. If RANCID were to be liberated, its name would need to change -- "Really Awesome" (or perhaps just the "Awesome") would need to be upgraded to something better. Someone would need to compile a list of superlatives beginning with 'a' for the acronym to stay the same. That'd be the first hurdle. cheers, Dale -------------- next part -------------- An HTML attachment was scrubbed... URL: From Dan.Weintraub at dealer.com Mon Feb 8 19:02:34 2016 From: Dan.Weintraub at dealer.com (Dan Weintraub) Date: Mon, 8 Feb 2016 19:02:34 +0000 Subject: [rancid] rancid git setup question In-Reply-To: <20160206010942.GA73515@shrubbery.net> References: <98193ABC-A665-4E5D-B7AB-BCEBA29136EA@dealer.com> <20160206010942.GA73515@shrubbery.net> Message-ID: On 2/5/16, 20:09 , "heasley" wrote: >I need to make that clearer; no attempt will be made to create repos if >CVSROOT looks like a URL. the script does not handle remote repos. >ie: if you want the origin (if i have the parlance correct) to be on a >remote host, you must create it yourself. My suggestion would be that you >let rancid-cvs create the repo locally, then move that to your remote host >and update group/.git/config to reflect that. > >CVSROOT does not translate well to git; for rancid's purpose, it puts all >of it origins - one per group - under CVSROOT. so CVSROOT/myGroup, >CVSROOT/youGroup, CVSROOT/groupies etc. And, the working directory is >BASEDIR/myGroup. So, if I change the origin of BASEDIR/myGroup then CVSROOT/myGroup no longer serves any purpose? Thanks, Dan Dan Weintraub | Systems Engineer V: 877.327.8422 x mailto:Dan.Weintraub at dealer.com | www.dealer.com From heas at shrubbery.net Mon Feb 8 23:17:29 2016 From: heas at shrubbery.net (heasley) Date: Mon, 8 Feb 2016 23:17:29 +0000 Subject: [rancid] rancid git setup question In-Reply-To: References: <98193ABC-A665-4E5D-B7AB-BCEBA29136EA@dealer.com> <20160206010942.GA73515@shrubbery.net> Message-ID: <20160208231729.GB79817@shrubbery.net> Mon, Feb 08, 2016 at 07:02:34PM +0000, Dan Weintraub: > On 2/5/16, 20:09 , "heasley" wrote: > > > >I need to make that clearer; no attempt will be made to create repos if > >CVSROOT looks like a URL. the script does not handle remote repos. > >ie: if you want the origin (if i have the parlance correct) to be on a > >remote host, you must create it yourself. My suggestion would be that you > >let rancid-cvs create the repo locally, then move that to your remote host > >and update group/.git/config to reflect that. > > > >CVSROOT does not translate well to git; for rancid's purpose, it puts all > >of it origins - one per group - under CVSROOT. so CVSROOT/myGroup, > >CVSROOT/youGroup, CVSROOT/groupies etc. And, the working directory is > >BASEDIR/myGroup. > > So, if I change the origin of BASEDIR/myGroup then CVSROOT/myGroup no longer serves any purpose? correct; if you move CVSROOT/myGroup to your central repo server, then its "moved" and CVSROOT/myGroup is no longer needed. > Thanks, > Dan > > > > > > > Dan Weintraub | Systems Engineer > V: 877.327.8422 x > mailto:Dan.Weintraub at dealer.com | www.dealer.com > > > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From penguin-kernel at I-love.SAKURA.ne.jp Tue Feb 9 10:30:16 2016 From: penguin-kernel at I-love.SAKURA.ne.jp (Tetsuo Handa) Date: Tue, 9 Feb 2016 19:30:16 +0900 Subject: [rancid] Two suggestions about rancid-3.3.0 Message-ID: <201602091930.CCG95819.FQStJHOFOFLMVO@I-love.SAKURA.ne.jp> Hello. I just looked at rancid-3.3.0.tar.gz and noticed two bugs. Bug1: . is not removed from PATH in the following four files. /usr/local/rancid-3.3/etc/lg.conf $ENV{PATH}="/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin"; /usr/local/rancid-3.3/etc/rancid.conf PATH=/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin; export PATH /usr/local/rancid-3.3/share/rancid/lg.conf.sample $ENV{PATH}="/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin"; /usr/local/rancid-3.3/share/rancid/rancid.conf.sample PATH=/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin; export PATH Bug2: @PACKAGE@ and @VERSION@ are not replaced in the following three files. /usr/local/rancid-3.3/bin/clogin send_user "@PACKAGE@ @VERSION@\n" /usr/local/rancid-3.3/bin/jlogin send_user "@PACKAGE@ @VERSION@\n" /usr/local/rancid-3.3/bin/mtlogin send_user "@PACKAGE@ @VERSION@\n" Regards. From david.ahrens at citrix.com Tue Feb 9 15:45:17 2016 From: david.ahrens at citrix.com (David Ahrens) Date: Tue, 9 Feb 2016 15:45:17 +0000 Subject: [rancid] Brocade 200 switches not working with Rancid Message-ID: <642FAF39868FD54A909CBFC1DD8F61A70D6E81EE@SJCPEX01CL02.citrite.net> Hi, I have a number of old Brocade 200 SAN switches that I'm trying to add to rancid so I can have rancid backup the switch configs. I added the following lines to rancid.types.base for Brocade: # # Brocade 200a # brocade;script;brocaderancid brocade;login;brocadelogin Here's my router.db rancid-test.eng.citrite.net;smc;up cc3750g;cisco;up brocade200;brocade;up I've attached the brocadelogin and brocaderancid scripts for collecting the switch configs. When I run rancid-run, the attached log files shows the switch configuration was collected from the switch, but no data was written to brocade200.raw. The file brocade200 contains: !RANCID-CONTENT-TYPE: Brocade ! ! I have rancid working with my Cisco and Dell PowerConnect switches, so I know rancid is installed and configured properly. Thanks for your help, David -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: brocadelogin Type: application/octet-stream Size: 26803 bytes Desc: brocadelogin URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: brocaderancid Type: application/octet-stream Size: 16418 bytes Desc: brocaderancid URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: EI.20160209.074020 Type: application/octet-stream Size: 25516 bytes Desc: EI.20160209.074020 URL: From david.ahrens at citrix.com Wed Feb 10 23:52:10 2016 From: david.ahrens at citrix.com (David Ahrens) Date: Wed, 10 Feb 2016 23:52:10 +0000 Subject: [rancid] Brocade 200 switches not working with Rancid Message-ID: <642FAF39868FD54A909CBFC1DD8F61A70D6EA92E@SJCPEX01CL02.citrite.net> I haven't see any responses to this post. Any idea why rancid would write output to the log file but not to brocade.raw? Thanks, David From: David Ahrens Sent: Tuesday, February 09, 2016 7:45 AM To: rancid-discuss at shrubbery.net Subject: Brocade 200 switches not working with Rancid Hi, I have a number of old Brocade 200 SAN switches that I'm trying to add to rancid so I can have rancid backup the switch configs. I added the following lines to rancid.types.base for Brocade: # # Brocade 200a # brocade;script;brocaderancid brocade;login;brocadelogin Here's my router.db rancid-test.eng.citrite.net;smc;up cc3750g;cisco;up brocade200;brocade;up I've attached the brocadelogin and brocaderancid scripts for collecting the switch configs. When I run rancid-run, the attached log files shows the switch configuration was collected from the switch, but no data was written to brocade200.raw. The file brocade200 contains: !RANCID-CONTENT-TYPE: Brocade ! ! I have rancid working with my Cisco and Dell PowerConnect switches, so I know rancid is installed and configured properly. Thanks for your help, David -------------- next part -------------- An HTML attachment was scrubbed... URL: From missakiiki at gmail.com Thu Feb 11 09:27:52 2016 From: missakiiki at gmail.com (Sharone B.) Date: Thu, 11 Feb 2016 01:27:52 -0800 (PST) Subject: [rancid] SSR SmartEdge/Ericsson Router - rancid compatibility? Message-ID: Hi Guys, I'm sorry if this has been posted before, I was wondering whether these anyone has managed to get rancid working with these devices. When I try to login via clogin I get the error below; spawn hpuifilter -- ssh -c 3des -x -l ranciduser X.X.X.X no matching cipher found: client 3des-cbc server aes128-ctr,aes192-ctr, aes256-ctr,arcfour256,aes256-cbc,arcfour128 Error: Couldn't login . Any advice is welcome. ~Sharone -------------- next part -------------- An HTML attachment was scrubbed... URL: From stony at stony.com Mon Feb 15 09:56:34 2016 From: stony at stony.com (Viktor Steinmann) Date: Mon, 15 Feb 2016 10:56:34 +0100 Subject: [rancid] HP 1920 Message-ID: <56C1A0D2.2000804@stony.com> Hey all I'd like to take up the discussion from June 2015 about HP 1920 switches. > http://www.shrubbery.net/pipermail/rancid-discuss/2015-June/008557.html I have purchased some of these switches as well, and in fact the Web interface and the CLI suck. But you can get to a full CLI with some extra commands: a) Enter "_cmdline-mode on" b) The switch answers with "All commands can be displayed and executed. Continue? [Y/N]" on which you answer with "Y" c) The switch will answer with the password prompt: "Please input password:" d) The password to enter here is "Jinhua1920unauthorized" e) The switch will output this: "Warning: Now you enter an all-command mode for developer's testing, some commands may affect operation by wrong use, please carefully use it with our engineer's direction." f) Now you enter "system-view", which gives you the "enable" mode on this switch-type Now you have full access to the CLI, which is very much standard (e.g. with "display current-configuration" you'll get the full configuration). Hope this helps. Kind regards, Viktor From jethro.binks at strath.ac.uk Mon Feb 15 10:05:15 2016 From: jethro.binks at strath.ac.uk (Jethro R Binks) Date: Mon, 15 Feb 2016 10:05:15 +0000 (GMT) Subject: [rancid] HP 1920 In-Reply-To: <56C1A0D2.2000804@stony.com> References: <56C1A0D2.2000804@stony.com> Message-ID: On Mon, 15 Feb 2016, Viktor Steinmann wrote: > I'd like to take up the discussion from June 2015 about HP 1920 switches. > > > http://www.shrubbery.net/pipermail/rancid-discuss/2015-June/008557.html > > I have purchased some of these switches as well, and in fact the Web interface > and the CLI suck. But you can get to a full CLI with some extra commands: > > a) Enter "_cmdline-mode on" > b) The switch answers with "All commands can be displayed and executed. > Continue? [Y/N]" on which you answer with "Y" > c) The switch will answer with the password prompt: "Please input password:" > d) The password to enter here is "Jinhua1920unauthorized" That's a new one. It used to be "512900" on the 1910s (at least). I haven't had any time to review my hacks to get this handled by Rancid 3 I'm afraid :( . Jethro. . . . . . . . . . . . . . . . . . . . . . . . . . Jethro R Binks, Network Manager, Information Services Directorate, University Of Strathclyde, Glasgow, UK The University of Strathclyde is a charitable body, registered in Scotland, number SC015263. From heas at shrubbery.net Mon Feb 15 12:16:40 2016 From: heas at shrubbery.net (heasley) Date: Mon, 15 Feb 2016 12:16:40 +0000 Subject: [rancid] HP 1920 In-Reply-To: References: <56C1A0D2.2000804@stony.com> Message-ID: <20160215121640.GI80648@shrubbery.net> Mon, Feb 15, 2016 at 10:05:15AM +0000, Jethro R Binks: > On Mon, 15 Feb 2016, Viktor Steinmann wrote: > > > I'd like to take up the discussion from June 2015 about HP 1920 switches. > > > > > http://www.shrubbery.net/pipermail/rancid-discuss/2015-June/008557.html > > > > I have purchased some of these switches as well, and in fact the Web interface > > and the CLI suck. But you can get to a full CLI with some extra commands: > > > > a) Enter "_cmdline-mode on" > > b) The switch answers with "All commands can be displayed and executed. > > Continue? [Y/N]" on which you answer with "Y" > > c) The switch will answer with the password prompt: "Please input password:" > > d) The password to enter here is "Jinhua1920unauthorized" > > That's a new one. It used to be "512900" on the 1910s (at least). this sounds ugly. someone would have to provide me access to one in order for me to help with adding support - if its reasonable once in "full cli" mode. > I haven't had any time to review my hacks to get this handled by Rancid 3 > I'm afraid :( . > > Jethro. > > . . . . . . . . . . . . . . . . . . . . . . . . . > Jethro R Binks, Network Manager, > Information Services Directorate, University Of Strathclyde, Glasgow, UK > > The University of Strathclyde is a charitable body, registered in > Scotland, number SC015263. > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From heas at shrubbery.net Mon Feb 15 17:21:00 2016 From: heas at shrubbery.net (heasley) Date: Mon, 15 Feb 2016 17:21:00 +0000 Subject: [rancid] Two suggestions about rancid-3.3.0 In-Reply-To: <201602091930.CCG95819.FQStJHOFOFLMVO@I-love.SAKURA.ne.jp> References: <201602091930.CCG95819.FQStJHOFOFLMVO@I-love.SAKURA.ne.jp> Message-ID: <20160215172100.GD87878@shrubbery.net> Tue, Feb 09, 2016 at 07:30:16PM +0900, Tetsuo Handa: > Hello. > > I just looked at rancid-3.3.0.tar.gz and noticed two bugs. > > Bug1: . is not removed from PATH in the following four files. > > /usr/local/rancid-3.3/etc/lg.conf > $ENV{PATH}="/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin"; > > /usr/local/rancid-3.3/etc/rancid.conf > PATH=/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin; export PATH > > /usr/local/rancid-3.3/share/rancid/lg.conf.sample > $ENV{PATH}="/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin"; > > /usr/local/rancid-3.3/share/rancid/rancid.conf.sample > PATH=/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin; export PATH It properly filters the path for me. Please send the config.log to me. > Bug2: @PACKAGE@ and @VERSION@ are not replaced in the following three files. > > /usr/local/rancid-3.3/bin/clogin > send_user "@PACKAGE@ @VERSION@\n" > > /usr/local/rancid-3.3/bin/jlogin > send_user "@PACKAGE@ @VERSION@\n" > > /usr/local/rancid-3.3/bin/mtlogin > send_user "@PACKAGE@ @VERSION@\n" yes, I have a bug there. Another user supplied a fix and I'll drop 3.4 in a bit to address this. From lists at sumpfralle.de Mon Feb 15 19:30:19 2016 From: lists at sumpfralle.de (Lars Kruse) Date: Mon, 15 Feb 2016 20:30:19 +0100 Subject: [rancid] Gathering switch information via SNMP or web interfaces Message-ID: <20160215203019.767bfbcb@erker.lan> Hello, currently I am involved with the maintenance of a dozen switches (Netgear GS724T, Netgear GS728TP and Netgear GS748T). I would like to distribute notifications to our small group of administrators after a switch configuration change was made by one of us. As far as I understand RANCID, I assume that it could be used for this purpose. Sadly our switches do not allow commandline access (ssh or telnet). They offer only a web interface and SNMP. I took a quick look at the rancid code and came to the conclusion that it uses expect for all its operations. Thus I could imagine that it only works with switches with commandline interfaces. Please correct my assumptions above, if necessary. I am prepared to write some code (extending RANCID or writing single-purpose scripts for our environment). But I cannot judge if it is possible (and desirable) to extend RANCID in a way to would allow the gathering of information via SNMP or web interfaces (with all the related complications of sessions and/or cookies). Would someone please share some thoughts regarding this approach? Thank you for your time! Cheers, Lars From howie at thingy.com Mon Feb 15 21:34:43 2016 From: howie at thingy.com (Howard Jones) Date: Mon, 15 Feb 2016 21:34:43 +0000 Subject: [rancid] Gathering switch information via SNMP or web interfaces In-Reply-To: <20160215203019.767bfbcb@erker.lan> References: <20160215203019.767bfbcb@erker.lan> Message-ID: <56C24473.4040209@thingy.com> On 15/02/2016 19:30, Lars Kruse wrote: > Hello, > > currently I am involved with the maintenance of a dozen switches (Netgear > GS724T, Netgear GS728TP and Netgear GS748T). > I would like to distribute notifications to our small group of administrators > after a switch configuration change was made by one of us. > As far as I understand RANCID, I assume that it could be used for this purpose. > > Sadly our switches do not allow commandline access (ssh or telnet). They offer > only a web interface and SNMP. > > I took a quick look at the rancid code and came to the conclusion that it uses > expect for all its operations. Thus I could imagine that it only works with > switches with commandline interfaces. > > Please correct my assumptions above, if necessary. > > I am prepared to write some code (extending RANCID or writing single-purpose > scripts for our environment). > But I cannot judge if it is possible (and desirable) to extend RANCID in a way > to would allow the gathering of information via SNMP or web interfaces (with all > the related complications of sessions and/or cookies). > Would someone please share some thoughts regarding this approach? > > Thank you for your time! > I used to have some old Nortel/Bay switches that had no CLI. I used wraprancid, which allows you to use the output of a script as input to RANCID, and then wrote a small perl program that polled the switches with SNMP and output a 'cisco-like' config for them, just covering port descriptions, speeds and VLAN membership (they were just 24-port 10/100 access switches). That allowed me to use them with RANCID, and also meant that I could drop in 2950s as replacements when they went pop :-) Interface info comes from the standard IF-MIB and IP-MIB. VLAN membership is either a vendor-specific thing or via Q-BRIDGE-MIB. Someone on StackExchange is using Q-BRIDGE-MIB with Netgear, so that should do it for you. The very basic script could literally be a few snmpwalk commands, and leave RANCID to spot the diffs. http://serverfault.com/questions/441273/configure-vlan-on-netgear-switch-via-snmp Howard From heas at shrubbery.net Mon Feb 15 21:46:11 2016 From: heas at shrubbery.net (heasley) Date: Mon, 15 Feb 2016 21:46:11 +0000 Subject: [rancid] Gathering switch information via SNMP or web interfaces In-Reply-To: <20160215203019.767bfbcb@erker.lan> References: <20160215203019.767bfbcb@erker.lan> Message-ID: <20160215214610.GB96658@shrubbery.net> Mon, Feb 15, 2016 at 08:30:19PM +0100, Lars Kruse: > Hello, > > currently I am involved with the maintenance of a dozen switches (Netgear > GS724T, Netgear GS728TP and Netgear GS748T). > I would like to distribute notifications to our small group of administrators > after a switch configuration change was made by one of us. > As far as I understand RANCID, I assume that it could be used for this purpose. > > Sadly our switches do not allow commandline access (ssh or telnet). They offer > only a web interface and SNMP. > > I took a quick look at the rancid code and came to the conclusion that it uses > expect for all its operations. Thus I could imagine that it only works with > switches with commandline interfaces. > > Please correct my assumptions above, if necessary. > > I am prepared to write some code (extending RANCID or writing single-purpose > scripts for our environment). > But I cannot judge if it is possible (and desirable) to extend RANCID in a way > to would allow the gathering of information via SNMP or web interfaces (with all > the related complications of sessions and/or cookies). > Would someone please share some thoughts regarding this approach? Yes, I would like to add something to do this, preferably maintaining the current organization or process so that it is not disjoint. I have just not found the time to work on it. > Thank you for your time! > > Cheers, > Lars > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From penguin-kernel at I-love.SAKURA.ne.jp Tue Feb 16 10:03:17 2016 From: penguin-kernel at I-love.SAKURA.ne.jp (Tetsuo Handa) Date: Tue, 16 Feb 2016 19:03:17 +0900 Subject: [rancid] Two suggestions about rancid-3.3.0 In-Reply-To: <20160215172100.GD87878@shrubbery.net> References: <201602091930.CCG95819.FQStJHOFOFLMVO@I-love.SAKURA.ne.jp> <20160215172100.GD87878@shrubbery.net> Message-ID: <201602161903.HHC09329.OFJOFSFVLtHMQO@I-love.SAKURA.ne.jp> heasley wrote: > Tue, Feb 09, 2016 at 07:30:16PM +0900, Tetsuo Handa: > > Hello. > > > > I just looked at rancid-3.3.0.tar.gz and noticed two bugs. > > > > Bug1: . is not removed from PATH in the following four files. > > > > /usr/local/rancid-3.3/etc/lg.conf > > $ENV{PATH}="/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin"; > > > > /usr/local/rancid-3.3/etc/rancid.conf > > PATH=/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin; export PATH > > > > /usr/local/rancid-3.3/share/rancid/lg.conf.sample > > $ENV{PATH}="/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin"; > > > > /usr/local/rancid-3.3/share/rancid/rancid.conf.sample > > PATH=/usr/local/rancid-3.3/bin:/usr/bin:/usr/sbin:/bin:.:/usr/local/bin:/usr/bin; export PATH > > It properly filters the path for me. Please send the config.log to me. > My environment does not have rsh and telnet. ---------- # grep ac_cv_path_ config.log ac_cv_path_COMM=/usr/bin/comm ac_cv_path_CVS=/usr/bin/cvs ac_cv_path_DIRNAME=/usr/bin/dirname ac_cv_path_EGREP='/bin/grep -E' ac_cv_path_EXPECT_PATH=/usr/bin/expect ac_cv_path_FIND=/bin/find ac_cv_path_GREP=/bin/grep ac_cv_path_ID=/usr/bin/id ac_cv_path_MAKE=/usr/bin/gmake ac_cv_path_MKDIR=/bin/mkdir ac_cv_path_MKTEMP_PATH=/bin/mktemp ac_cv_path_PERLV_PATH=/usr/bin/perl ac_cv_path_PING_PATH=/bin/ping ac_cv_path_RSH=no ac_cv_path_SENDMAIL=/usr/sbin/sendmail ac_cv_path_SORT=/bin/sort ac_cv_path_SSH=/usr/bin/ssh ac_cv_path_TELNET=no ac_cv_path_TOUCH=/bin/touch ac_cv_path_install='/usr/bin/install -c' ac_cv_path_mkdir=/bin/mkdir ---------- Passing "no" to dirname command returns ".". ---------- # dirname no . ---------- Therefore, unconditional concatenation results in "." in ENV_PATH. ENV_PATH="$ENV_PATH:`dirname $ac_cv_path_RSH`" ENV_PATH="$ENV_PATH:`dirname $ac_cv_path_TELNET`" Like I suggested on Fri, 5 Jun 2015 20:45:38 +0900, you can later filter "." from ENV_PATH. From heas at shrubbery.net Tue Feb 16 11:05:24 2016 From: heas at shrubbery.net (heasley) Date: Tue, 16 Feb 2016 11:05:24 +0000 Subject: [rancid] Two suggestions about rancid-3.3.0 In-Reply-To: <201602161903.HHC09329.OFJOFSFVLtHMQO@I-love.SAKURA.ne.jp> References: <201602091930.CCG95819.FQStJHOFOFLMVO@I-love.SAKURA.ne.jp> <20160215172100.GD87878@shrubbery.net> <201602161903.HHC09329.OFJOFSFVLtHMQO@I-love.SAKURA.ne.jp> Message-ID: <20160216110524.GA19362@shrubbery.net> Tue, Feb 16, 2016 at 07:03:17PM +0900, Tetsuo Handa: > you can later filter "." from ENV_PATH. It was, but not properly. the fix will in the next release. From stony at stony.com Tue Feb 16 11:39:30 2016 From: stony at stony.com (Viktor Steinmann) Date: Tue, 16 Feb 2016 12:39:30 +0100 Subject: [rancid] HP 1920 In-Reply-To: <20160215121640.GI80648@shrubbery.net> References: <56C1A0D2.2000804@stony.com> <20160215121640.GI80648@shrubbery.net> Message-ID: <56C30A72.10201@stony.com> On 15.02.2016 13:16, heasley wrote: > this sounds ugly. someone would have to provide me access to one in > order for me to help with adding support - if its reasonable once in > "full cli" mode. It is indeed ugly. I already have the switches in production, so I can't give access, unfortunately. Sorry. Viktor From jethro.binks at strath.ac.uk Tue Feb 16 14:08:13 2016 From: jethro.binks at strath.ac.uk (Jethro R Binks) Date: Tue, 16 Feb 2016 14:08:13 +0000 (GMT) Subject: [rancid] HP 1920 In-Reply-To: <56C30A72.10201@stony.com> References: <56C1A0D2.2000804@stony.com> <20160215121640.GI80648@shrubbery.net> <56C30A72.10201@stony.com> Message-ID: On Tue, 16 Feb 2016, Viktor Steinmann wrote: > On 15.02.2016 13:16, heasley wrote: > > this sounds ugly. someone would have to provide me access to one in order > > for me to help with adding support - if its reasonable once in "full cli" > > mode. > It is indeed ugly. I already have the switches in production, so I can't give > access, unfortunately. All the grunt work is there for running commands and interpreting output for Comware (3/4/7), incorporated into the comware support I have. However extra commands require to be issued for the models that initially have this limited CLI to get to the full CLI. In order to know to execute them, extra information needs to be provided for the cmwlogin script, and I think it was that part that required me to do a local hack, and that's the part that my memory is most hazy on. It needs to be recalled, extracted, and fed back as a patch (assuming there isn't a neater way of doing it) against something rather more up to date than the rancid3 version I'm currently running. Jethro. . . . . . . . . . . . . . . . . . . . . . . . . . Jethro R Binks, Network Manager, Information Services Directorate, University Of Strathclyde, Glasgow, UK The University of Strathclyde is a charitable body, registered in Scotland, number SC015263. From incudie at gmail.com Wed Feb 17 00:43:27 2016 From: incudie at gmail.com (Incudie) Date: Tue, 16 Feb 2016 16:43:27 -0800 Subject: [rancid] Gathering switch information via SNMP or web interfaces In-Reply-To: <20160215203019.767bfbcb@erker.lan> References: <20160215203019.767bfbcb@erker.lan> Message-ID: I've been working on something similar for Sonicwalls using python and selenium. Perhaps you can have selenium call up Firefox into xvfb (to keep it headless), connect to your device and download the config, then pipe the config into rancid using the wraprancid script mentioned. Hopefully I'm understanding your correctly. -Tim On Feb 15, 2016 11:55 AM, "Lars Kruse" wrote: > Hello, > > currently I am involved with the maintenance of a dozen switches (Netgear > GS724T, Netgear GS728TP and Netgear GS748T). > I would like to distribute notifications to our small group of > administrators > after a switch configuration change was made by one of us. > As far as I understand RANCID, I assume that it could be used for this > purpose. > > Sadly our switches do not allow commandline access (ssh or telnet). They > offer > only a web interface and SNMP. > > I took a quick look at the rancid code and came to the conclusion that it > uses > expect for all its operations. Thus I could imagine that it only works with > switches with commandline interfaces. > > Please correct my assumptions above, if necessary. > > I am prepared to write some code (extending RANCID or writing > single-purpose > scripts for our environment). > But I cannot judge if it is possible (and desirable) to extend RANCID in a > way > to would allow the gathering of information via SNMP or web interfaces > (with all > the related complications of sessions and/or cookies). > Would someone please share some thoughts regarding this approach? > > Thank you for your time! > > Cheers, > Lars > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss > -------------- next part -------------- An HTML attachment was scrubbed... URL: From heas at shrubbery.net Wed Feb 17 10:21:27 2016 From: heas at shrubbery.net (John Heasley) Date: Wed, 17 Feb 2016 10:21:27 +0000 Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> References: <20160127072712.GC33037@shrubbery.net> <56A8967A.9000007@thingy.com> <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> Message-ID: <20160217102127.GA52532@shrubbery.net> Wed, Feb 03, 2016 at 06:37:45PM +0000, David Ahrens: > Hi Elliot, > > I have some old Brocade 200 SAN switches running version 5.2.2 of Fabric OS. The brocaderancid script seems to run properly but for some reason doesn't recognize/capture the out of the commands. The attached log file shows the output of the switch commands, any ideas? So, how many different O/S's does brocade have? FabricOS, what else? I've cleaned-up the brocadelogin script a little bit and imported it as brlogin. This could probably be cleaned-up further, remove ExtremeOS-specific stuff etc, but I have nothing to test against. ftp://ftp.shrubbery.net/pub/rancid/alpha/rancid-3.4.99.tar.gz > Thanks, > David > > > > -----Original Message----- > From: Rancid-discuss [mailto:rancid-discuss-bounces at shrubbery.net] On Behalf Of Elliot Wilen > Sent: Friday, January 29, 2016 11:00 AM > To: rancid-discuss at shrubbery.net > Subject: Re: [rancid] rancid support for Brocade Fabric OS > > Comparing Howard's version of the script to mine, I see a few differences, one of which may be a modification I made. Instead of using configShow directly, my script uses configUpload -all and saves it to a local file on the switch, followed by configShow -l to display that file. I think this yields a more complete configuration similar to what one would use to back up the switch via ftp. > > His version also has a couple additions relative to whatever the common source was. > > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From jethro.binks at strath.ac.uk Wed Feb 17 10:40:28 2016 From: jethro.binks at strath.ac.uk (Jethro R Binks) Date: Wed, 17 Feb 2016 10:40:28 +0000 (GMT) Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <20160217102127.GA52532@shrubbery.net> References: <20160127072712.GC33037@shrubbery.net> <56A8967A.9000007@thingy.com> <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> <20160217102127.GA52532@shrubbery.net> Message-ID: FabricOS is the SAN stuff. Then there's the network device OS, which is mostly supported in the foundry/flogin (Brocade bought Foundry years ago). It varies a bit across devices, at the moment two similar-but-different code bases are being integrated, but it is superficially the same and config file is cisco-ish. I think if brocadelogin/etc is being written for the SAN product line FabricOS, it should be explicit about that, as otherwise there will be confusion to anyone coming to Brocade networking in the last several years, selecting this new brocadelogin, and not realising they actually want the Foundry stuff. Call it brocadesanlogin etc? Jethro. On Wed, 17 Feb 2016, John Heasley wrote: > Wed, Feb 03, 2016 at 06:37:45PM +0000, David Ahrens: > > Hi Elliot, > > > > I have some old Brocade 200 SAN switches running version 5.2.2 of Fabric OS. The brocaderancid script seems to run properly but for some reason doesn't recognize/capture the out of the commands. The attached log file shows the output of the switch commands, any ideas? > > So, how many different O/S's does brocade have? FabricOS, what else? > > I've cleaned-up the brocadelogin script a little bit and imported it as > brlogin. This could probably be cleaned-up further, remove ExtremeOS-specific > stuff etc, but I have nothing to test against. > > ftp://ftp.shrubbery.net/pub/rancid/alpha/rancid-3.4.99.tar.gz > > > Thanks, > > David > > > > > > > > -----Original Message----- > > From: Rancid-discuss [mailto:rancid-discuss-bounces at shrubbery.net] On Behalf Of Elliot Wilen > > Sent: Friday, January 29, 2016 11:00 AM > > To: rancid-discuss at shrubbery.net > > Subject: Re: [rancid] rancid support for Brocade Fabric OS > > > > Comparing Howard's version of the script to mine, I see a few differences, one of which may be a modification I made. Instead of using configShow directly, my script uses configUpload -all and saves it to a local file on the switch, followed by configShow -l to display that file. I think this yields a more complete configuration similar to what one would use to back up the switch via ftp. > > > > His version also has a couple additions relative to whatever the common source was. > > > > _______________________________________________ > > Rancid-discuss mailing list > > Rancid-discuss at shrubbery.net > > http://www.shrubbery.net/mailman/listinfo/rancid-discuss > > > > > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss > . . . . . . . . . . . . . . . . . . . . . . . . . Jethro R Binks, Network Manager, Information Services Directorate, University Of Strathclyde, Glasgow, UK The University of Strathclyde is a charitable body, registered in Scotland, number SC015263. From howie at thingy.com Wed Feb 17 10:43:05 2016 From: howie at thingy.com (Howard Jones) Date: Wed, 17 Feb 2016 10:43:05 +0000 Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <20160217102127.GA52532@shrubbery.net> References: <20160127072712.GC33037@shrubbery.net> <56A8967A.9000007@thingy.com> <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> <20160217102127.GA52532@shrubbery.net> Message-ID: <56C44EB9.4080402@thingy.com> On 17/02/2016 10:21, John Heasley wrote: > Wed, Feb 03, 2016 at 06:37:45PM +0000, David Ahrens: >> Hi Elliot, >> >> I have some old Brocade 200 SAN switches running version 5.2.2 of Fabric OS. The brocaderancid script seems to run properly but for some reason doesn't recognize/capture the out of the commands. The attached log file shows the output of the switch commands, any ideas? > So, how many different O/S's does brocade have? FabricOS, what else? > > I've cleaned-up the brocadelogin script a little bit and imported it as > brlogin. This could probably be cleaned-up further, remove ExtremeOS-specific > stuff etc, but I have nothing to test against. > > ftp://ftp.shrubbery.net/pub/rancid/alpha/rancid-3.4.99.tar.gz > Brocade bought Foundry, so they have FabricOS/FabOS for their own traditional SAN Switches (silkworm etc), plus the Foundry OS (Ironware?) which is a lot like ios. They also recently bought Vyatta, which has its own config style. IMHO vendor names are not a good choice for device types - 'cisco' is really 'ios', not css/arrowpoint, or nxos, or iosxr, or catos. Those are all supported, but sometimes only if you know which vendor bought which 10 years ago... same sort of thing for companies like Dell and HP where it's not even different ranges with different OS - some HP is comware and some is procurve-classic, all in the Procurve range, as far as I can see. Juniper SSG is another - 'netscreen' isn't in the product name or company name anymore. From heas at shrubbery.net Wed Feb 17 13:04:10 2016 From: heas at shrubbery.net (heasley) Date: Wed, 17 Feb 2016 13:04:10 +0000 Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <56C44EB9.4080402@thingy.com> References: <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> <20160217102127.GA52532@shrubbery.net> <56C44EB9.4080402@thingy.com> Message-ID: <20160217130410.GA55823@shrubbery.net> Wed, Feb 17, 2016 at 10:43:05AM +0000, Howard Jones: > Brocade bought Foundry, so they have FabricOS/FabOS for their own > traditional SAN Switches (silkworm etc), plus the Foundry OS (Ironware?) > which is a lot like ios. They also recently bought Vyatta, which has its > own config style. > > IMHO vendor names are not a good choice for device types - 'cisco' is > really 'ios', not css/arrowpoint, or nxos, or iosxr, or catos. Those are > all supported, but sometimes only if you know which vendor bought which > 10 years ago... same sort of thing for companies like Dell and HP where > it's not even different ranges with different OS - some HP is comware > and some is procurve-classic, all in the Procurve range, as far as I can > see. Juniper SSG is another - 'netscreen' isn't in the product name or > company name anymore. i'm happy to rename the foundry scripts and convert them to the rancid 3.0-style; but someone will have to help test as I no longer have this equipment. the brocadelogin script folks sent to me looks like a mix of flogin and clogin. If i've coalesced all the comments together, that script works for fabricos & ironware (at least some versions thereof). Do I have this right? but the brocaderancid script is for fabricos, given the command set? {'chassisShow' => 'ChassisShow'}, {'firmwareShow' => 'FirmwareShow'}, {'version' => 'FirmwareShow'}, {'zoneshow' => 'FirmwareShow'}, {'portcfgshow' => 'FirmwareShow'}, #{'fosConfig --show' => 'Fosconfig'}, {'ipAddrShow' => 'IpAddrShow'}, {'licenseShow' => 'LicenseShow'}, {'domainsShow' => 'DomainsShow'}, {'configupload -force -all -l config.txt' => 'ConfigShow'}, {'configShow -l config.txt' => 'ConfigShow'} and, i received scripts from another user today that I have not looked through yet. From allan at physics.umn.edu Wed Feb 17 15:05:09 2016 From: allan at physics.umn.edu (Graham Allan) Date: Wed, 17 Feb 2016 09:05:09 -0600 Subject: [rancid] Gathering switch information via SNMP or web interfaces In-Reply-To: References: <20160215203019.767bfbcb@erker.lan> Message-ID: <56C48C25.1070908@physics.umn.edu> Possibly something like perl's WWW::Mechanize would be simpler for this. Graham On 2/16/2016 6:43 PM, Incudie wrote: > I've been working on something similar for Sonicwalls using python and > selenium. Perhaps you can have selenium call up Firefox into xvfb (to > keep it headless), connect to your device and download the config, then > pipe the config into rancid using the wraprancid script mentioned. > > Hopefully I'm understanding your correctly. > > -Tim From jason at biel-tech.com Wed Feb 17 17:32:16 2016 From: jason at biel-tech.com (Jason Biel) Date: Wed, 17 Feb 2016 12:32:16 -0500 Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <20160217130410.GA55823@shrubbery.net> References: <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> <20160217102127.GA52532@shrubbery.net> <56C44EB9.4080402@thingy.com> <20160217130410.GA55823@shrubbery.net> Message-ID: <-6681253939640117142@unknownmsgid> *plug* github would be perfect for things like this *plug* > On Feb 17, 2016, at 08:04, heasley wrote: > > Wed, Feb 17, 2016 at 10:43:05AM +0000, Howard Jones: >> Brocade bought Foundry, so they have FabricOS/FabOS for their own >> traditional SAN Switches (silkworm etc), plus the Foundry OS (Ironware?) >> which is a lot like ios. They also recently bought Vyatta, which has its >> own config style. >> >> IMHO vendor names are not a good choice for device types - 'cisco' is >> really 'ios', not css/arrowpoint, or nxos, or iosxr, or catos. Those are >> all supported, but sometimes only if you know which vendor bought which >> 10 years ago... same sort of thing for companies like Dell and HP where >> it's not even different ranges with different OS - some HP is comware >> and some is procurve-classic, all in the Procurve range, as far as I can >> see. Juniper SSG is another - 'netscreen' isn't in the product name or >> company name anymore. > > i'm happy to rename the foundry scripts and convert them to the rancid > 3.0-style; but someone will have to help test as I no longer have this > equipment. > > the brocadelogin script folks sent to me looks like a mix of flogin and > clogin. If i've coalesced all the comments together, that script works > for fabricos & ironware (at least some versions thereof). Do I have this > right? > > but the brocaderancid script is for fabricos, given the command set? > > {'chassisShow' => 'ChassisShow'}, > {'firmwareShow' => 'FirmwareShow'}, > {'version' => 'FirmwareShow'}, > {'zoneshow' => 'FirmwareShow'}, > {'portcfgshow' => 'FirmwareShow'}, > #{'fosConfig --show' => 'Fosconfig'}, > {'ipAddrShow' => 'IpAddrShow'}, > {'licenseShow' => 'LicenseShow'}, > {'domainsShow' => 'DomainsShow'}, > {'configupload -force -all -l config.txt' => 'ConfigShow'}, > {'configShow -l config.txt' => 'ConfigShow'} > > and, i received scripts from another user today that I have not looked > through yet. > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From nick at foobar.org Wed Feb 17 17:33:20 2016 From: nick at foobar.org (Nick Hilliard) Date: Wed, 17 Feb 2016 17:33:20 +0000 Subject: [rancid] rancid support for Brocade Fabric OS In-Reply-To: <-6681253939640117142@unknownmsgid> References: <20160128004808.GH63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F883FA@ATC1.corp.aechelon.com> <20160128013022.GL63800@shrubbery.net> <736C3AA4541A56418A535328BDC357590100F98F8A@ATC1.corp.aechelon.com> <20160128193339.GJ1520@shrubbery.net> <56AB3B79.7030108@thingy.com> <736C3AA4541A56418A535328BDC357590100FA1354@ATC1.corp.aechelon.com> <642FAF39868FD54A909CBFC1DD8F61A70D6DF3C1@SJCPEX01CL02.citrite.net> <20160217102127.GA52532@shrubbery.net> <56C44EB9.4080402@thingy.com> <20160217130410.GA55823@shrubbery.net> <-6681253939640117142@unknownmsgid> Message-ID: <56C4AEE0.1090000@foobar.org> Jason Biel wrote: > *plug* github would be perfect for things like this *plug* yeah, it really really would. Pull requests make community workflow way easier. Nick From serge.torop at gmail.com Thu Feb 18 20:51:10 2016 From: serge.torop at gmail.com (Serge Torop) Date: Thu, 18 Feb 2016 23:51:10 +0300 Subject: [rancid] Rancid can't fully save the config file of D-Link switch Message-ID: Hello, all, I have installed v 3.3.0 and try to work with D-Link switches. My configuration scripts is: 1. etc/rancid.types.base: dlink1;script;rancid -t dell dlink1;login;dllogin dlink1;module;dell dlink1;inloop;dell::inloop dlink1;command;dlink1::GetSystem;show switch dlink1;command;dlink1::GetConf;show config effective 2. router.db: 10.37.11.210;dlink1;up 10.37.4.184;dlink1;up I have run "rancid-run dlinkdevices" and see config files /usr/local/rancid/var/dlinkdevices/config/10.37.11.210.new and 10.37.4.184.new with normal config content. After script is finished in log file I see the messages: Trying to get all of the configs. 10.37.11.210: End of run not found #------------------------------------------------------------------- 10.37.4.184: End of run not found #------------------------------------------------------------------- ===================================== Getting missed routers: round 1. 10.37.11.210: End of run not found #------------------------------------------------------------------- 10.37.4.184: End of run not found And config files (10.37.11.210 and 10.37.4.184) are empty or config information not comletely saved (with some parts of configuration). Test command, for example, ./dllogin -c "show config" 10.37.4.184 is work fine. Can you help me find the misconfiguration or other working method? Thank you in advance. -- Serge P. Torop From heas at shrubbery.net Thu Feb 18 21:16:05 2016 From: heas at shrubbery.net (heasley) Date: Thu, 18 Feb 2016 21:16:05 +0000 Subject: [rancid] Rancid can't fully save the config file of D-Link switch In-Reply-To: References: Message-ID: <20160218211605.GD97307@shrubbery.net> Thu, Feb 18, 2016 at 11:51:10PM +0300, Serge Torop: > Hello, all, > > I have installed v 3.3.0 and try to work with D-Link switches. > > My configuration scripts is: > > 1. > etc/rancid.types.base: you should put your custom stuff in rancid.types.conf > dlink1;script;rancid -t dell > dlink1;login;dllogin > dlink1;module;dell > dlink1;inloop;dell::inloop > dlink1;command;dlink1::GetSystem;show switch > dlink1;command;dlink1::GetConf;show config effective some dells are SMC OEMs; see the smc type. > 2. > router.db: > > 10.37.11.210;dlink1;up > 10.37.4.184;dlink1;up > > I have run "rancid-run dlinkdevices" and see config files > /usr/local/rancid/var/dlinkdevices/config/10.37.11.210.new and 10.37.4.184.new > with normal config content. > > After script is finished in log file I see the messages: > > Trying to get all of the configs. > 10.37.11.210: End of run not found > #------------------------------------------------------------------- > 10.37.4.184: End of run not found > #------------------------------------------------------------------- > ===================================== > Getting missed routers: round 1. > 10.37.11.210: End of run not found > #------------------------------------------------------------------- > 10.37.4.184: End of run not found > > And config files (10.37.11.210 and 10.37.4.184) are empty or config > information not comletely saved > (with some parts of configuration). > > Test command, for example, ./dllogin -c "show config" 10.37.4.184 is work fine. > > Can you help me find the misconfiguration or other working method? > > Thank you in advance. > > > -- > Serge P. Torop > _______________________________________________ > Rancid-discuss mailing list > Rancid-discuss at shrubbery.net > http://www.shrubbery.net/mailman/listinfo/rancid-discuss From A.Hongens at netmatch.nl Sat Feb 20 15:53:50 2016 From: A.Hongens at netmatch.nl (=?iso-8859-1?Q?Angelo_H=F6ngens?=) Date: Sat, 20 Feb 2016 15:53:50 +0000 Subject: [rancid] problem with rancid on centos7 Message-ID: <6A7ABA19243F1E4EADD8BB1563CDDCCB8A77950E@TIL-EXCH-05.netmatch.local> Hey, I have an issue with rancid that I've been trying to debug. Rancid from epel won't work on my CentOS 7 boxes, and I'm hoping one of you can help me. I'm using the epel rpm, since I deploy everything from puppet (no custom compiling or whatever). I first thought it was a bug in the packaging by the epel package manager (and opened a bugzilla report), but I found out that if I recompile the package myself I get the same result. I just blew away my VM and installed a new CentOS Linux release 7.1.1503 machine from a clean template (I use for >200 other machines). I did a yum install rancid, which installed rancid-3.2-2.el7.x86_64 from epel7. Why I see is that clogin works as root (and gives a correct error): [root at nmo-ran-02 ~]# /usr/libexec/rancid/clogin foo Error: password file (/root/.cloginrc) does not exist But not as user: [rancid at nmo-ran-02 ~]$ /usr/libexec/rancid/clogin foo Error: invalid command name "aadd" I have no idea where this aadd command comes from.. I see no differences in bash profile configs or the like. Ideas? Angelo Hongens From shawn at rmrf.us Sat Feb 20 15:15:29 2016 From: shawn at rmrf.us (Shawn Laemmrich) Date: Sat, 20 Feb 2016 10:15:29 -0500 Subject: [rancid] Rancid failing after OS upgrade Message-ID: I've had rancid 3.2 running (on ubuntu) for weeks now without issues. Yesterday, I upgraded the OS (ubuntu do-release-upgrade) and now rancid is giving me cvs errors when I run rancid-run property 'svn:ignore' set on '.' property 'svn:ignore' set on 'configs' Updating '.': At revision 3. Sending . Sending configs Sending router.db Transmitting file data . Committed revision 4. svn: E125001: '/var/lib/rancid/bob/configs/.cvsignore' does not exist Deleted .cvsignore I usually use CVS instead of SVN but thought I'd give it a shot, since I got similar errors with CVS. When I run rancid-run, I get this error in the log file almost immediately and it puts all my hosts in routers.down So far I've tried verifying clogin and all that stuff (works fine, I can use clogin to get into anything) removing my rancid config, and re-running rancid-cvs to create the required directories (didn't help) removing and purging rancid and cvs from the machine and re-installing (verified it installed correctly) running rancid by hand rancid -t cisco xxx.yyy.zzz.aaa for a device, which works fine and creates a device.new file correctly. So, it definitely seems to be a CVS thing, but for the life of me I can't figure out what it's mad about or how to make it work again. -------------- next part -------------- An HTML attachment was scrubbed... URL: From penguin-kernel at I-love.SAKURA.ne.jp Sun Feb 21 03:55:49 2016 From: penguin-kernel at I-love.SAKURA.ne.jp (Tetsuo Handa) Date: Sun, 21 Feb 2016 12:55:49 +0900 Subject: [rancid] problem with rancid on centos7 In-Reply-To: <6A7ABA19243F1E4EADD8BB1563CDDCCB8A77950E@TIL-EXCH-05.netmatch.local> References: <6A7ABA19243F1E4EADD8BB1563CDDCCB8A77950E@TIL-EXCH-05.netmatch.local> Message-ID: <201602211255.ACF18245.OLSFFVMtQJOFHO@I-love.SAKURA.ne.jp> Angelo Hongens wrote: > Why I see is that clogin works as root (and gives a correct error): > > [root at nmo-ran-02 ~]# /usr/libexec/rancid/clogin foo > Error: password file (/root/.cloginrc) does not exist This is expected behavior because environment variable HOME is set to /root for user root . There is no ~root/.cloginrc . > > But not as user: > > [rancid at nmo-ran-02 ~]$ /usr/libexec/rancid/clogin foo > Error: invalid command name "aadd" This is because you have a typo in ~rancid/.cloginrc . I think that command meant to be "add" not "aadd". From a.hongens at netmatch.nl Sun Feb 21 12:51:19 2016 From: a.hongens at netmatch.nl (=?UTF-8?Q?Angelo_H=c3=b6ngens?=) Date: Sun, 21 Feb 2016 13:51:19 +0100 Subject: [rancid] problem with rancid on centos7 In-Reply-To: <201602211255.ACF18245.OLSFFVMtQJOFHO@I-love.SAKURA.ne.jp> References: <6A7ABA19243F1E4EADD8BB1563CDDCCB8A77950E@TIL-EXCH-05.netmatch.local> <201602211255.ACF18245.OLSFFVMtQJOFHO@I-love.SAKURA.ne.jp> Message-ID: <56C9B2C7.9080703@netmatch.nl> On 2/21/2016 4:55 AM, Tetsuo Handa wrote: >> [rancid at nmo-ran-02 ~]$ /usr/libexec/rancid/clogin foo >> Error: invalid command name "aadd" > > This is because you have a typo in ~rancid/.cloginrc . > I think that command meant to be "add" not "aadd". > *facepalm* Thanks guys, you're right. It was typo in .cloginrc. Angelo. From josem.junk at gmail.com Tue Feb 23 20:23:44 2016 From: josem.junk at gmail.com (Jose Martinez) Date: Tue, 23 Feb 2016 14:23:44 -0600 Subject: [rancid] .cvsignore files showing in viewvc Message-ID: Hello all, I was using an old RANCID version 2.X on an Ubuntu server. I'm creating a new server using the newest rancid. I have everything working but when I look on the web interface of ViewVC i see .cvsignore files along with the configs. Anyway to hide these or anything like that? I don't see these files being created with the old RANCID installation. -------------- next part -------------- An HTML attachment was scrubbed... URL: From heas at shrubbery.net Tue Feb 23 22:18:42 2016 From: heas at shrubbery.net (heasley) Date: Tue, 23 Feb 2016 22:18:42 +0000 Subject: [rancid] .cvsignore files showing in viewvc In-Reply-To: References: Message-ID: <20160223221842.GC94395@shrubbery.net> Tue, Feb 23, 2016 at 02:23:44PM -0600, Jose Martinez: > Hello all, > > I was using an old RANCID version 2.X on an Ubuntu server. > > I'm creating a new server using the newest rancid. I have everything > working but when I look on the web interface of ViewVC i see .cvsignore > files along with the configs. > > Anyway to hide these or anything like that? I don't see these files being > created with the old RANCID installation. They just tell the SCM to ignore certain files. Which means that those will not appear in cvs update, etc opertions, which means less confusion in logs for the less experienced users. if it taints your view too much, alter viewvc to filter those files. From josem.junk at gmail.com Tue Feb 23 22:37:45 2016 From: josem.junk at gmail.com (Jose Martinez) Date: Tue, 23 Feb 2016 16:37:45 -0600 Subject: [rancid] .cvsignore files showing in viewvc In-Reply-To: <20160223221842.GC94395@shrubbery.net> References: <20160223221842.GC94395@shrubbery.net> Message-ID: Thanks Heasley. I'm trying to research exactly what you mean. I'm a novice at Linux. I'm not sure what SCM is. I've also tried researching "viewvc filters" but not finding much about that either. Any push in the right direction would be appreciated. On Tue, Feb 23, 2016 at 4:18 PM, heasley wrote: > Tue, Feb 23, 2016 at 02:23:44PM -0600, Jose Martinez: > > Hello all, > > > > I was using an old RANCID version 2.X on an Ubuntu server. > > > > I'm creating a new server using the newest rancid. I have everything > > working but when I look on the web interface of ViewVC i see .cvsignore > > files along with the configs. > > > > Anyway to hide these or anything like that? I don't see these files being > > created with the old RANCID installation. > > They just tell the SCM to ignore certain files. Which means that those > will not appear in cvs update, etc opertions, which means less confusion > in logs for the less experienced users. > > if it taints your view too much, alter viewvc to filter those files. > -------------- next part -------------- An HTML attachment was scrubbed... URL: From chris at ceegeebee.com Sat Feb 27 02:58:50 2016 From: chris at ceegeebee.com (Chris Bennett) Date: Sat, 27 Feb 2016 13:28:50 +1030 Subject: [rancid] Revision controlled full backup without notification? Message-ID: <20160227025850.GA3984@cgb-linux.rcmb.lan> Hi there, Is there a way of using RANCID to perform standard config backups with password removal etc & change notification, coupled with a full untouched backup of the config in revision control without change notification? I'm wanting to ensure I have both useful change notification emails, whilst having a full copy of the config for restore/DR purposes. Thanks, Chris From ler762 at gmail.com Sat Feb 27 05:51:39 2016 From: ler762 at gmail.com (Lee) Date: Sat, 27 Feb 2016 00:51:39 -0500 Subject: [rancid] Revision controlled full backup without notification? In-Reply-To: <20160227025850.GA3984@cgb-linux.rcmb.lan> References: <20160227025850.GA3984@cgb-linux.rcmb.lan> Message-ID: On 2/26/16, Chris Bennett wrote: > Hi there, > > Is there a way of using RANCID to perform standard config backups with > password removal etc & change notification, coupled with a full > untouched backup of the config in revision control without change > notification? I'm wanting to ensure I have both useful change > notification emails, whilst having a full copy of the config for > restore/DR purposes. It's pretty easy - control_rancid puts the diff output into $TMP.diff So have rancid do unmodified config backups & add something like the following # -LR- begin: remove passwords etc. from diff listing sanitize.sh $TMP.diff >$TMP.diff2 /bin/mv $TMP.diff2 $TMP.diff in control_rancid to remove the sensitive bits from the file before mailing it out. Regards, Lee $ cat sanitize.sh #!/bin/sh # script to sanitize a ciscso config awk ' /^Index: configs/ { if ( NR != 1) printf("\n\n\n") } { # IOS stuph strip("key 7 ", "xXxXxX") strip("md5 7 ", "xXxXxX") strip("password 3 ", "xXxXxX") strip("password 5 ", "xXxXxX") strip("password 7 ", "xXxXxX") strip("secret 5 ", "xXxXxX") strip("snmp-server community ", "xXxXxX") strip("crypto isakmp key ", "xXxXxX") strip("tacacs-server key 7 ", "xXxXxX") # CatOS stuph strip("set password " , "xXxXxX") strip("set enablepass ", "xXxXxX") strip("set snmp community read-only " , "xXxXxX") strip("set snmp community read-write " , "xXxXxX") strip("set snmp community read-write-all ", "xXxXxX") strip("set tacacs key ", "xXxXxX") print $0 } func strip(s, r) { sub(s " *[^ ]*", s r, $0) } ' $1 $ From chris at ceegeebee.com Sat Feb 27 06:56:36 2016 From: chris at ceegeebee.com (Chris Bennett) Date: Sat, 27 Feb 2016 17:26:36 +1030 Subject: [rancid] Revision controlled full backup without notification? In-Reply-To: References: <20160227025850.GA3984@cgb-linux.rcmb.lan> Message-ID: <20160227065636.GB3984@cgb-linux.rcmb.lan> Hey thanks for the reply. What you propose will work great for one use case I have. I was actually thinking of another use case, but now that I've looked closer at what RANCID does, it doesn't look possible without some coding. I'm dealing with a couple of devices that that have cycling RSA keys. The XXrancid script specifically excludes these snippets unconditionally. If the config fetching was separate from the processing of output, I could do what I want easily :) Thanks, Chris From heas at shrubbery.net Sat Feb 27 15:26:02 2016 From: heas at shrubbery.net (heasley) Date: Sat, 27 Feb 2016 15:26:02 +0000 Subject: [rancid] Revision controlled full backup without notification? In-Reply-To: References: <20160227025850.GA3984@cgb-linux.rcmb.lan> Message-ID: <20160227152602.GC19969@shrubbery.net> Sat, Feb 27, 2016 at 12:51:39AM -0500, Lee: > It's pretty easy - control_rancid puts the diff output into $TMP.diff > So have rancid do unmodified config backups & add something like the following > > # -LR- begin: remove passwords etc. from diff listing > sanitize.sh $TMP.diff >$TMP.diff2 > /bin/mv $TMP.diff2 $TMP.diff I encourage you to use rancid.conf:DIFFSCRIPT added in rancid 3.2 to include your custom filter. Otherwise, the right solution. Prost. From heas at shrubbery.net Sat Feb 27 15:29:22 2016 From: heas at shrubbery.net (heasley) Date: Sat, 27 Feb 2016 15:29:22 +0000 Subject: [rancid] Revision controlled full backup without notification? In-Reply-To: <20160227065636.GB3984@cgb-linux.rcmb.lan> References: <20160227025850.GA3984@cgb-linux.rcmb.lan> <20160227065636.GB3984@cgb-linux.rcmb.lan> Message-ID: <20160227152922.GD19969@shrubbery.net> Sat, Feb 27, 2016 at 05:26:36PM +1030, Chris Bennett: > I was actually thinking of another use case, but now that I've looked > closer at what RANCID does, it doesn't look possible without some > coding. I'm dealing with a couple of devices that that have cycling > RSA keys. The XXrancid script specifically excludes these snippets > unconditionally. If the config fetching was separate from the > processing of output, I could do what I want easily :) I'd be willing to change the code to optionally retain things like this, but still filter other oscillating things like timestamps. It potentially makes the history rather cluttered. But, perhaps someone has a better idea to address this?