From JasonT at plumtree.com Thu Oct 16 17:36:51 2003 From: JasonT at plumtree.com (Jason Truong) Date: Thu, 16 Oct 2003 10:36:51 -0700 Subject: how use an edited version of cat5rancid Message-ID: Hello, I have successfully installed rancid and it working great for some of my Cisco switches. The ones that work are for the Cisco Catalyst 4003. I have a bunch of Catalyst 3548s that don't appear to work. In my router.db file I have: cat4003-mez.plumtree.com:cat5:up cat4003-2nd.plumtree.com:cat5:up cat4003-3rd.plumtree.com:cat5:up cat4006-server.plumtree.com:cat5:up cat3548-2nd.plumtree.com:cat5:up cat3548-3rd.plumtree.com:cat5:up cat3548-svr-sw1.plumtree.com:cat5:up cat3548-svr-sw2.plumtree.com:cat5:up cat3548-svr-sw3.plumtree.com:cat5:up cat3548-svr-sw4.plumtree.com:cat5:up So I am using "cat5" for mfg for my cisco catalyst switches. However, when I run do-diffs, nothing happens for the cat3548s. When I run rancid or cat5rancid (for debugging) I get errors. Basically, some of the called commands from the scripts don't work .. ie [rancid at Nagios bin]$ ./cat5rancid -d cat3548-2nd.plumtree.com executing clogin -t 90 -c"show version;show boot;show flash;dir bootflash:;dir slot0:;dir slot1:;dir sup-bootflash:;dir sup-microcode:;show module;show port ifindex;write term" cat3548-2nd.plumtree.com cat3548-2nd.plumtree.com: missed cmd(s): show port ifindex,show module,dir sup-microcode:,dir sup-bootflash:,dir bootflash:,dir slot0:,show version,show flash,write term,show boot,dir slot1: cat3548-2nd.plumtree.com: missed cmd(s): show port ifindex,show module,dir sup-microcode:,dir sup-bootflash:,dir bootflash:,dir slot0:,show version,show flash,write term,show boot,dir slot1: cat3548-2nd.plumtree.com: End of run not found cat3548-2nd.plumtree.com: End of run not found Some of the commands mention above dont' work. It appears that cat5rancid does not output anything if any of the commands don't work while rancid at least spits out whatever it can find. So my questions are: - how do we know if rancid or cat5rancid is called? (should I use cisco as the mfg in my router.db file?) - how do we get rancid to call an edited version of any of these scripts? thanks, Jason T. From sid at tsnz.net Thu Oct 16 19:29:54 2003 From: sid at tsnz.net (Sid Jones) Date: Fri, 17 Oct 2003 08:29:54 +1300 (NZDT) Subject: Session 1 on a 2900 In-Reply-To: References: Message-ID: Has anybody already done anything to connect to [for example] an ATM daughter card in 2900 and scrape the config of that off as well as the runnig config on the base switch? From asp at partan.com Thu Oct 16 19:41:05 2003 From: asp at partan.com (Andrew Partan) Date: Thu, 16 Oct 2003 15:41:05 -0400 Subject: how use an edited version of cat5rancid In-Reply-To: References: Message-ID: <20031016194105.GA53829@partan.com> On Thu, Oct 16, 2003 at 10:36:51AM -0700, Jason Truong wrote: > So I am using "cat5" for mfg for my cisco catalyst switches. > However, when I run do-diffs, nothing happens for the cat3548s. > When I run rancid or cat5rancid (for debugging) I get errors. > Basically, some of the called commands from the scripts don't work I never remember which ciscos are best done with type 'cisco' (rancid) or type 'cat5' (cat5rancid). I usually use rancid 1st, and if that fails, then use cat5rancid. [See bin/rancid-fe for the bit that goes from type to program.] In any case, the 'missed cmds' message means that the program (cat5rancid in this case) has problems logging into the router & could not run all of the commands. The programs deal with 'command not found' or 'invalid command' just fine; its when they can't even run the commands that they complain about missed commands. So I'd look into why you can't log into the router & run commands. Can you use clogin to log into the routers having problems? Can you run some commands with: clogin -c "show version;show version" rtr1 without problems? Btw, the missed commands that you listed are *all* of the commands that cat5rancid is trying to run - i.e.: it was unable to run any of them. --asp From krueger.b at portseattle.org Thu Oct 16 19:44:39 2003 From: krueger.b at portseattle.org (Krueger, Brian) Date: Thu, 16 Oct 2003 12:44:39 -0700 Subject: Special Characters Message-ID: Has anybody run into problems when using special characters with passwords? When I try to run .cloginrc it gives me this (password characters have been changed): Bareword found where operator expected at .cloginrc line 70, near "$ add autoenable" (Missing operator before autoenable?) Number found where operator expected at .cloginrc line 70, near "org 01" (Do you need to predeclare org?) Semicolon seems to be missing at .cloginrc line 70. Operator or semicolon missing before *step at .cloginrc line 118. Ambiguous use of * resolved as operator *step at .cloginrc line 118. syntax error at .cloginrc line 70, near "$ add autoenable P69" Execution of .cloginrc aborted due to compilation errors. Any help would be greatly appreciated BK krueger.b at portseattle.org From JSahala at fusiontel.com Thu Oct 16 19:47:36 2003 From: JSahala at fusiontel.com (Joshua Sahala) Date: Thu, 16 Oct 2003 15:47:36 -0400 Subject: Special Characters Message-ID: you have to escape the '$' character, so if your password is $3 References: Message-ID: <01L1WHD5MCPO000NR6@tmk.com> > I never remember which ciscos are best done with type 'cisco' > (rancid) or type 'cat5' (cat5rancid). I usually use rancid 1st, > and if that fails, then use cat5rancid. [See bin/rancid-fe for the > bit that goes from type to program.] If it runs IOS, use "cisco". If it runs CatOS, use "cat5". I probably should have put a check in cat5rancid to see if the target device was running IOS and return a "you're using the wrong device type" error, but cat5rancid was originally a hack for my own use only. A switch running in hybrid mode should be treated like a C5K with an attached RSM - separate router.db entries, one of type "cat5" for the switch and one of type "cisco" for the router. Native mode switches only need a single "cisco" entry for the whole switch/router combo. Similarly, any module that you get to with the "session" command is running a private copy of IOS and should be configured as a separate entry in router.db of type "cisco". Terry Kennedy http://www.tmk.com terry at tmk.com New York, NY USA From asp at partan.com Thu Oct 16 19:50:55 2003 From: asp at partan.com (Andrew Partan) Date: Thu, 16 Oct 2003 15:50:55 -0400 Subject: Special Characters In-Reply-To: References: Message-ID: <20031016195055.GA57590@partan.com> On Thu, Oct 16, 2003 at 03:47:36PM -0400, Joshua Sahala wrote: > you have to escape the '$' character, so if your password is $3 you would need to put \$3 References: Message-ID: >I have successfully installed rancid and it working great for some >of my Cisco switches. >The ones that work are for the Cisco Catalyst 4003. > >I have a bunch of Catalyst 3548s that don't appear to work. >In my router.db file I have: > >So I am using "cat5" for mfg for my cisco catalyst switches. >However, when I run do-diffs, nothing happens for the cat3548s. When >I run rancid or cat5rancid (for debugging) I get errors. Basically, >some of the called commands from the scripts don't work .. ie Catalyst 3548's run IOS so you need to use the 'cisco' RANCID directive. On the other hand, Catalyst 2980's (which are Catalyst 4000 series switches) run CatOS. So just because the model is a Catalyst doesn't mean it is running CatOS. I don't know why the Cisco naming is so convoluted. If a 'show ver' on the switch reveals that the binary OS image starts with 'cat' such as cat4000.6-3-3a.bin then it's running CatOS and you need to use the RANCID directive 'cat5'. Otherwise it is running IOS and needs to use RANCID directive 'cisco'. There may be exceptions to this. That is the way it looks to me based on the Cisco equipment I happen to have. Mark From yuvalba at netvision.net.il Fri Oct 17 11:55:11 2003 From: yuvalba at netvision.net.il (Yuval Ben-Ari) Date: Fri, 17 Oct 2003 13:55:11 +0200 Subject: false "changes in XXX routers" reports Message-ID: <070F6E0C1E5E684BAA2DC8C854386D4A01CE4E@ntx.forest.netvision.net.il> Hi, I'm getting flase reports of changes in router.db file, such as: Routers changed to up: router1:cisco router2:cisco router3:cisco Added routers: router1:cisco router2:cisco router3:cisco Deleted routers: router1:cisco router2:cisco router3:cisco while there was no change to the routers list file. any idea why this happens? --Yuval From afort at choqolat.org Sun Oct 19 22:40:08 2003 From: afort at choqolat.org (Andrew Fort) Date: Mon, 20 Oct 2003 08:40:08 +1000 Subject: Session 1 on a 2900 In-Reply-To: References: Message-ID: <3F9312C8.8000809@choqolat.org> Sid Jones wrote: > Has anybody already done anything to connect to [for example] an ATM > daughter card in 2900 and scrape the config of that off as well as the > runnig config on the base switch? Not that I'm aware of. If it's just a show command, it's easy to add, otherwise there's a bit more work. If you like, can you show us what the exchange to get that config looks like (from global exec mode on the main switch IOS) (and also, what the 'hostname' line of your switch is)? cheers -afort From asp at partan.com Mon Oct 20 16:14:02 2003 From: asp at partan.com (Andrew Partan) Date: Mon, 20 Oct 2003 12:14:02 -0400 Subject: false "changes in XXX routers" reports In-Reply-To: <070F6E0C1E5E684BAA2DC8C854386D4A01CE4E@ntx.forest.netvision.net.il> References: <070F6E0C1E5E684BAA2DC8C854386D4A01CE4E@ntx.forest.netvision.net.il> Message-ID: <20031020161402.GA28286@partan.com> On Fri, Oct 17, 2003 at 01:55:11PM +0200, Yuval Ben-Ari wrote: > I'm getting flase reports of changes in router.db file, such as: Rancid remembers which routers were up or down in rancid/$GROUP/routers.up & .down. If this file is being deleted or can't be written to or created by the user running rancid, you might get these problems. --asp From krueger.b at portseattle.org Mon Oct 20 20:44:47 2003 From: krueger.b at portseattle.org (Krueger, Brian) Date: Mon, 20 Oct 2003 13:44:47 -0700 Subject: Special Characters Message-ID: I think I'm doing something wrong. I added {} to .cloginrc and ran it, it still gave me output saying: [root at p69rh186 rancid]# perl .cloginrc Array found where operator expected at .cloginrc line 69, at end of line Semicolon seems to be missing at .cloginrc line 69. Number found where operator expected at .cloginrc line 70, near "Core1 1" (Do you need to predeclare Core1?) Semicolon seems to be missing at .cloginrc line 70. Operator or semicolon missing before &step at .cloginrc line 71. Ambiguous use of & resolved as operator & at .cloginrc line 71. syntax error at .cloginrc line 70, near "r0und@} " Missing right curly or square bracket at .cloginrc line 118, at end of line Execution of .cloginrc aborted due to compilation errors. Here's my .cloginrc script (at least it's all that's not commented out) add user P69-Core1 jb3 add password P69-Core1 {r0und@} add autoenable P69-Core1 1 add password t91-wanrtr {l00k&step} From asp at partan.com Mon Oct 20 21:09:17 2003 From: asp at partan.com (Andrew Partan) Date: Mon, 20 Oct 2003 17:09:17 -0400 Subject: Special Characters In-Reply-To: References: Message-ID: <20031020210917.GA49144@partan.com> On Mon, Oct 20, 2003 at 01:44:47PM -0700, Krueger, Brian wrote: > [root at p69rh186 rancid]# perl .cloginrc Its not a perl program. Its an expect script sourced by the *login scripts. --asp From krueger.b at portseattle.org Mon Oct 20 21:16:11 2003 From: krueger.b at portseattle.org (Krueger, Brian) Date: Mon, 20 Oct 2003 14:16:11 -0700 Subject: Special Characters Message-ID: Ok, so how would I test it? From asp at partan.com Mon Oct 20 21:22:40 2003 From: asp at partan.com (Andrew Partan) Date: Mon, 20 Oct 2003 17:22:40 -0400 Subject: Special Characters In-Reply-To: References: Message-ID: <20031020212239.GA49379@partan.com> On Mon, Oct 20, 2003 at 02:16:11PM -0700, Krueger, Brian wrote: > Ok, so how would I test it? Try logging into the router with the appropriate *login script. e.g.: clogin P69-Core1 clogin t91-wanrtr [or jlogin for junipers or ...] --asp From sid at tsnz.net Mon Oct 20 22:36:53 2003 From: sid at tsnz.net (Sid Jones) Date: Tue, 21 Oct 2003 11:36:53 +1300 (NZDT) Subject: Session 1 on a 2900 In-Reply-To: <3F9312C8.8000809@choqolat.org> References: <3F9312C8.8000809@choqolat.org> Message-ID: On Mon, 20 Oct 2003, Andrew Fort wrote: > Sid Jones wrote: > > Has anybody already done anything to connect to [for example] an ATM > > daughter card in 2900 and scrape the config of that off as well as the > > runnig config on the base switch? > > Not that I'm aware of. If it's just a show command, it's easy to add, > otherwise there's a bit more work. If you like, can you show us what > the exchange to get that config looks like (from global exec mode on the > main switch IOS) (and also, what the 'hostname' line of your switch is)? switch>enable Password: switch#session 1 Entering Console for module in slot 1 Type "^G" to end this session Press RETURN to get started! switch.ATM>en switch.ATM# switch.ATM#exit Disconnecting from slot 1 Connection Duration: 00:01:36 switch# From steve at smiller.org Mon Oct 20 23:26:56 2003 From: steve at smiller.org (Stephen Miller) Date: Mon, 20 Oct 2003 16:26:56 -0700 Subject: Session 1 on a 2900 In-Reply-To: References: <3F9312C8.8000809@choqolat.org> Message-ID: <20031020231213.M82041@smiller.org> of course it's doable...but the question is how far to dig into it. if you just need to start gathering the daughter card configs right away...then you can copy and modify a login and rancid scripts to just log into the daughter cards on each device. You would need to create a new group for them....but since there can be more than one daughter card per device...you would need to make the rancid script iterate through all the daughter cards on the device...this would be a bit more complex. if you want the full blown, gather all the card configs and put them under each device...keeping with standard CVS structure...then this would take much more effort than the one above. Here would be issues with password of the daughter card in cloginrc, checking the total number of cards in the device, etc...this could be another sub within rancid.. just throwing this out there...to see what other options there are. Modifying a login and rancid script to just gather the configs from a daughter card should be pretty straight-forward. But trying to iterate through more than one card on the device...not sure...would have to take a closer look at rancid since that is where it should be done at. sorry not more help...just talking out loud ;-) steve ---------- Original Message ----------- From: Sid Jones To: Andrew Fort Cc: Sid Jones , rancid-discuss at shrubbery.net Sent: Tue, 21 Oct 2003 11:36:53 +1300 (NZDT) Subject: Re: Session 1 on a 2900 > On Mon, 20 Oct 2003, Andrew Fort wrote: > > > Sid Jones wrote: > > > Has anybody already done anything to connect to [for example] an ATM > > > daughter card in 2900 and scrape the config of that off as well as the > > > runnig config on the base switch? > > > > Not that I'm aware of. If it's just a show command, it's easy to add, > > otherwise there's a bit more work. If you like, can you show us what > > the exchange to get that config looks like (from global exec mode on the > > main switch IOS) (and also, what the 'hostname' line of your switch is)? > > switch>enable > Password: > switch#session 1 > Entering Console for module in slot 1 > Type "^G" to end this session > > Press RETURN to get started! > > switch.ATM>en > switch.ATM# > switch.ATM#exit > > Disconnecting from slot 1 > Connection Duration: 00:01:36 > > switch# ------- End of Original Message ------- From yuvalba at netvision.net.il Tue Oct 21 14:48:02 2003 From: yuvalba at netvision.net.il (Yuval Ben-Ari) Date: Tue, 21 Oct 2003 16:48:02 +0200 Subject: false "changes in XXX routers" reports Message-ID: <070F6E0C1E5E684BAA2DC8C854386D4A01CE8A@ntx.forest.netvision.net.il> thanks, rancid is run as cron job of user "rancid" whom has permissions to write to the files. also, the problem happens randomly and I can't find any pattern here. From heas at shrubbery.net Tue Oct 21 14:58:56 2003 From: heas at shrubbery.net (john heasley) Date: Tue, 21 Oct 2003 14:58:56 +0000 Subject: false "changes in XXX routers" reports In-Reply-To: <070F6E0C1E5E684BAA2DC8C854386D4A01CE8A@ntx.forest.netvision.net.il> References: <070F6E0C1E5E684BAA2DC8C854386D4A01CE8A@ntx.forest.netvision.net.il> Message-ID: <20031021145856.GA22610@shrubbery.net> Tue, Oct 21, 2003 at 04:48:02PM +0200, Yuval Ben-Ari: > thanks, > > rancid is run as cron job of user "rancid" whom has permissions to write > to the files. > also, the problem happens randomly and I can't find any pattern here. iirc, when control_rancid runs, it actually uses temporary files which are renamed to routers.{up,down}. so, the directory must also be writable. your safest bet/fix is likely; # cd ~rancid # chown -R rancid . From yuvalba at netvision.net.il Wed Oct 22 08:16:41 2003 From: yuvalba at netvision.net.il (Yuval Ben-Ari) Date: Wed, 22 Oct 2003 10:16:41 +0200 Subject: false "changes in XXX routers" reports Message-ID: <070F6E0C1E5E684BAA2DC8C854386D4A01CE92@ntx.forest.netvision.net.il> done will have to see ...... > -----Original Message----- > From: john heasley [mailto:heas at shrubbery.net] > Sent: Tuesday, October 21, 2003 4:59 PM > To: Yuval Ben-Ari > Cc: Andrew Partan; rancid-discuss at shrubbery.net > Subject: Re: false "changes in XXX routers" reports > > > Tue, Oct 21, 2003 at 04:48:02PM +0200, Yuval Ben-Ari: > > thanks, > > > > rancid is run as cron job of user "rancid" whom has > permissions to write > > to the files. > > also, the problem happens randomly and I can't find any > pattern here. > > iirc, when control_rancid runs, it actually uses temporary > files which are > renamed to routers.{up,down}. so, the directory must also be > writable. > > your safest bet/fix is likely; > # cd ~rancid > # chown -R rancid . From azheng at monterey.k12.ca.us Wed Oct 29 00:48:08 2003 From: azheng at monterey.k12.ca.us (Alex Zheng) Date: Tue, 28 Oct 2003 16:48:08 -0800 Subject: how to retrieve IOS image using RANCID? In-Reply-To: References: Message-ID: <6.0.0.22.0.20031028164706.03eb06f8@mail.monterey.k12.ca.us> Anyone used Rancid to retrieve IOS images from a router? And how if anyone knows? Thanks. Alex From heas at shrubbery.net Wed Oct 29 01:34:56 2003 From: heas at shrubbery.net (john heasley) Date: Tue, 28 Oct 2003 17:34:56 -0800 Subject: how to retrieve IOS image using RANCID? In-Reply-To: <6.0.0.22.0.20031028164706.03eb06f8@mail.monterey.k12.ca.us> References: <6.0.0.22.0.20031028164706.03eb06f8@mail.monterey.k12.ca.us> Message-ID: <20031029013456.GK6611@shrubbery.net> Tue, Oct 28, 2003 at 04:48:08PM -0800, Alex Zheng: > > Anyone used Rancid to retrieve IOS images from a router? And how if anyone > knows? > > Thanks. > > > Alex You;ll have to provide a better description of what you're trying to do. the two possibility that come to mind of what you might be trying are definitely possible. :) From steve at smiller.org Wed Oct 29 18:57:57 2003 From: steve at smiller.org (Stephen Miller) Date: Wed, 29 Oct 2003 11:57:57 -0700 Subject: how to retrieve IOS image using RANCID? In-Reply-To: <20031029013456.GK6611@shrubbery.net> References: <6.0.0.22.0.20031028164706.03eb06f8@mail.monterey.k12.ca.us> <20031029013456.GK6611@shrubbery.net> Message-ID: <20031029185338.M92101@smiller.org> do you want to add the IOS images to CVS? you should be able to...just not sure if you want to...the space could be large depending on how many devices and what the image size is, normally about 8+ megs now. You could just modify a login script and use the -c option with, "copy flash:image ftp://ftpserver", which should do it. Then just cron that....but you wouldn't get the change if one occurred...but you would still have the old and new images. steve ---------- Original Message ----------- From: john heasley To: Alex Zheng Cc: rancid-discuss at shrubbery.net Sent: Tue, 28 Oct 2003 17:34:56 -0800 Subject: Re: how to retrieve IOS image using RANCID? > Tue, Oct 28, 2003 at 04:48:08PM -0800, Alex Zheng: > > > > Anyone used Rancid to retrieve IOS images from a router? And how if anyone > > knows? > > > > Thanks. > > > > > > Alex > > You;ll have to provide a better description of what you're trying to do. > > the two possibility that come to mind of what you might be trying are > definitely possible. :) ------- End of Original Message ------- From heas at shrubbery.net Thu Oct 30 02:38:23 2003 From: heas at shrubbery.net (john heasley) Date: Wed, 29 Oct 2003 18:38:23 -0800 Subject: bug report---- missed cmd(s): show install active In-Reply-To: <3FA072FB.9060906@cisco.com> References: <3FA072FB.9060906@cisco.com> Message-ID: <20031030023823.GO13118@shrubbery.net> Wed, Oct 29, 2003 at 06:10:03PM -0800, Forier Zhang: > the rancid running perfect with my Cisco GSR routers, but when I > upgraded the IOS to some latest version like 12.0(26.3)S, the do-diffs > or 'rancid router-name' stopped with error message says: > > 10.21.28.4: missed cmd(s): show install active > > after some debug I found that the problem maybe came with the 'show > version' output format in the new IOS versions, most likely it's the > lines as below: > > WARNING: Old fab-loader in slot 0; use "upgrade fabric-downloader" to update > WARNING: Old MBus agent ROM in some slots; use "upgrade mbus-agent-rom" > to update > > looks the current version (2.2.2) can not handle the ';' correctly. > of course, the problem gone after I did the upgrading as indicated in > the above warning messages. (they gone from the 'show version') > > does anyone there see this issue ? > > thanks ! This is more than likely trigger not by the ';' or anything else in those messages, but by how 2.2.2 parsed them, where by we tried to also collect the "suggested fix" that followed these warning. At some point in the recent past, Cisco (that's you :) dropped the suggestion text. If the last warning in the list had been one for which rancid tried to collect the "suggested fix", then it could have inadvertantly eaten the next command-line, which in 2.2.2 would be 'show install active', and resulted in it believing that the command had not be run properly. Hence, we have the following poised for 2.3: ---------------------------- revision 1.143 date: 2003/09/10 19:33:58; author: heas; state: Exp; lines: +0 -8 drop the "suggested action" portion of GSR LC/RP ROM upgrade warnings. Cisco keeps changing the format. ---------------------------- warning, I do not know if this will cleanly apply to 2.2.2. It might have to be wedged in manually. Index: rancid.in =================================================================== RCS file: /home/rancid/.CVS/rancid/bin/rancid.in,v retrieving revision 1.142 retrieving revision 1.143 diff -u -r1.142 -r1.143 --- rancid.in 23 Aug 2003 19:10:32 -0000 1.142 +++ rancid.in 10 Sep 2003 19:33:58 -0000 1.143 @@ -296,14 +296,6 @@ ProcessHistory("COMMENTS","keysort","I0","!\n"); } ProcessHistory("COMMENTS","keysort","I1","! $_"); - if (/Recommend RP/i) { - # RP Slot Warning is only single line - $_ = "\n"; - } else { - # The line after the WARNING is what to do about it. - $_ = ; tr/\015//d; - } - ProcessHistory("COMMENTS","keysort","I1","! $_"); } if (/^Configuration register is (.*)$/) { $config_register=$1; From otter at surfnet.nl Thu Oct 30 21:19:09 2003 From: otter at surfnet.nl (Niels den Otter) Date: Thu, 30 Oct 2003 22:19:09 +0100 Subject: Rancid problem with 'WARNING' in 'show version' Message-ID: <20031030211908.GA6918@pangsit> Rancid (version 2.2.2) doesn't appear to handle lines that start with 'WARNING' during 'show version' well. Code says the following: # The line after the WARNING is what to do about it. $_ = ; tr/\015//d; ProcessHistory("COMMENTS","keysort","I1","! $_"); However on GSRs the warning is often given directly on the same line, eg. WARNING: Old MBus agent ROM in some slots; use "upgrade mbus-agent-rom" to update The $_ causes the next prompt line to be read and the next command 'show install active' never to be executed, which makes Rancid warn about unsuccesful attempts. Does anyone know if the warning really comes after the WARNING line sometimes? If so I guess an extra check is needed to see if the next line is the prompt line and to continue if so. -- Niels From asp at partan.com Thu Oct 30 21:37:14 2003 From: asp at partan.com (Andrew Partan) Date: Thu, 30 Oct 2003 16:37:14 -0500 Subject: Rancid problem with 'WARNING' in 'show version' In-Reply-To: <20031030211908.GA6918@pangsit> References: <20031030211908.GA6918@pangsit> Message-ID: <20031030213714.GA26869@partan.com> On Thu, Oct 30, 2003 at 10:19:09PM +0100, Niels den Otter wrote: > Rancid (version 2.2.2) doesn't appear to handle lines that start with > 'WARNING' during 'show version' well. Code says the following: > > # The line after the WARNING is what to do about it. > $_ = ; tr/\015//d; > ProcessHistory("COMMENTS","keysort","I1","! $_"); We removed this bit of code from the upcoming rancid 2.3 since cisco keeps mucking with it. See this message from yesterday on this same subject: From: john heasley To: Forier Zhang Cc: rancid-discuss at shrubbery.net Subject: Re: bug report---- missed cmd(s): show install active Message-ID: <20031030023823.GO13118 at shrubbery.net> --asp From otter at surfnet.nl Thu Oct 30 21:39:19 2003 From: otter at surfnet.nl (Niels den Otter) Date: Thu, 30 Oct 2003 22:39:19 +0100 Subject: Rancid problem with 'WARNING' in 'show version' In-Reply-To: <20031030213714.GA26869@partan.com> References: <20031030211908.GA6918@pangsit> <20031030213714.GA26869@partan.com> Message-ID: <20031030213919.GC5556@pangsit> Andrew, On Thursday, 30 October 2003, Andrew Partan wrote: > We removed this bit of code from the upcoming rancid 2.3 since cisco keeps > mucking with it. > > See this message from yesterday on this same subject: > From: john heasley > To: Forier Zhang > Cc: rancid-discuss at shrubbery.net > Subject: Re: bug report---- missed cmd(s): show install active > Message-ID: <20031030023823.GO13118 at shrubbery.net> I have missed that e-mail. Thanks for the quick response (and fix). -- Niels From heas at shrubbery.net Thu Oct 30 21:55:42 2003 From: heas at shrubbery.net (john heasley) Date: Thu, 30 Oct 2003 13:55:42 -0800 Subject: Rancid problem with 'WARNING' in 'show version' In-Reply-To: <20031030213919.GC5556@pangsit> References: <20031030211908.GA6918@pangsit> <20031030213714.GA26869@partan.com> <20031030213919.GC5556@pangsit> Message-ID: <20031030215542.GA25634@shrubbery.net> Thu, Oct 30, 2003 at 10:39:19PM +0100, Niels den Otter: > Andrew, > > On Thursday, 30 October 2003, Andrew Partan wrote: > > We removed this bit of code from the upcoming rancid 2.3 since cisco keeps > > mucking with it. > > > > See this message from yesterday on this same subject: > > From: john heasley > > To: Forier Zhang > > Cc: rancid-discuss at shrubbery.net > > Subject: Re: bug report---- missed cmd(s): show install active > > Message-ID: <20031030023823.GO13118 at shrubbery.net> > > I have missed that e-mail. Thanks for the quick response (and fix). > I created a patch for this which will apply cleanly to 2.2.2. hope this helps folks. ftp://ftp.shrubbery.net/pub/rancid/rancid-2.2.2.p3