From: G. F. <ga...@fr...> - 2008-06-05 12:13:23
|
Good day, Recently, I upgraded devmon from v0.2.2 to v0.3 final (devmon-0.3.0.tar.gz / 2008-04-03 06:52 on Sourceforce). Below are some hacks I had to do to make it working for me. Any comment is of course very welcome. 1) CONN test must be green All our firewalls have ICMP ping disabled. In our bb-hosts, they are all marked as "noping". So the "conn" status of all those firewalls is "clear". With v0.3, it looks that devmon is skipping all devices that are not marked as conn=green. This is not applicable for us. I guess we should have a global variable that gives all acceptable status (with default = green) but so far, I just changed the tests within dm_snmp.pm and dm_tests.pm to accept "clear" status. 2) UNPACK did not work In file modules/dm_tests.pm, function trans_unpack, line 782, changed : my @vars = unpack $type, $packed; to my @vars = unpack "$type*", $packed; (and the same at line 795) 3) SUBSTR did not work (btw, I have the feeling I already fixed this before...) In file modules/dm_tests.pm, function trans_substr, line 548 and following, changed: if(defined $length) { $oid_h->{'val'}{$leaf} = substr $string, $offset, $length; } else { $oid_h->{'val'}{$leaf} = substr $string, $offset; } to my $string_length = length($string); if ($offset <= $string_length) { if(defined $length) { if ($offset + $length <= $string_length) { $oid_h->{'val'}{$leaf} = substr $string, $offset, $length; } else { # Length outside string limit, review length $oid_h->{'val'}{$leaf} = substr $string, $offset, $string_length - $offset; } } else { $oid_h->{'val'}{$leaf} = substr $string, $offset; } } else { # Offset outside string limit, returns an empty string $oid_h->{'val'}{$leaf} = ''; } 4) VERSION is still "rc1" And a last very minor change : In modules/dm_config.pm, line 38 : changed version from 0.3.0-rc1 to 0.3.0 Have a great day. -- Gaëtan Frenoy <gaetan|à|frenoy.net> |
From: Buchan M. <bg...@st...> - 2008-06-06 15:32:56
|
On Thursday 05 June 2008 14:13:14 Gaëtan Frenoy wrote: > Good day, > > Recently, I upgraded devmon from v0.2.2 to v0.3 final > (devmon-0.3.0.tar.gz / 2008-04-03 06:52 on Sourceforce). > > Below are some hacks I had to do to make it working for me. > Any comment is of course very welcome. > > > 1) CONN test must be green > > All our firewalls have ICMP ping disabled. In our bb-hosts, they are > all marked as "noping". So the "conn" status of all those firewalls is > "clear". > > With v0.3, it looks that devmon is skipping all devices that are not > marked as conn=green. You can disable this feature globally by changing BBTYPE in devmon.cfg to something other than "hobbit". I see now this change was not documented in the CHANGELOG, but it was implemented before I took over maintainership. At present there is no way to disable it on a per-device basis. > This is not applicable for us. I guess we should have a global > variable that gives all acceptable status (with default = green) but so > far, I just changed the tests within dm_snmp.pm and dm_tests.pm to > accept "clear" status. > > > 2) UNPACK did not work > > In file modules/dm_tests.pm, function trans_unpack, line 782, changed : > > my @vars = unpack $type, $packed; > > to > > my @vars = unpack "$type*", $packed; > > (and the same at line 795) Hmm, it is working here, it is used in the compaq-server/log/transforms template file, and I have a number of servers using this template, and the values work correctly. Can you provide more detail on when/where/how it isn't working? > 3) SUBSTR did not work ... in specific cases. > (btw, I have the feeling I already fixed this before...) > > In file modules/dm_tests.pm, function trans_substr, line 548 and > following, changed: > > if(defined $length) { > $oid_h->{'val'}{$leaf} = substr $string, $offset, $length; > } > else { > $oid_h->{'val'}{$leaf} = substr $string, $offset; > } > > to > > my $string_length = length($string); > if ($offset <= $string_length) { > if(defined $length) { > if ($offset + $length <= $string_length) { > $oid_h->{'val'}{$leaf} = substr $string, $offset, > $length; } else { > # Length outside string limit, review length > $oid_h->{'val'}{$leaf} = substr $string, $offset, > $string_length - $offset; } > } else { > $oid_h->{'val'}{$leaf} = substr $string, $offset; > } > } else { > # Offset outside string limit, returns an empty string > $oid_h->{'val'}{$leaf} = ''; > } > > The fix looks fine. Could you send a diff instead? > 4) VERSION is still "rc1" > Yes, I noticed it a few weeks ago. > And a last very minor change : > > In modules/dm_config.pm, line 38 : changed version from 0.3.0-rc1 to > 0.3.0 I think what I will do is try and fix all the real bugs (no features) in the next week or three, and release a 0.3.1. Thanks for the feedback. Regards, Buchan |
From: G. F. <ga...@fr...> - 2008-06-30 08:24:01
Attachments:
dm_tests.pm.substr.diff
sonic-tz170.tar.bz2
|
Hello, Sorry for the delay. Definitely not enough hours in one day. Le Fri, 6 Jun 2008 17:32:41 +0200, Buchan Milne écrivait: > > 2) UNPACK did not work > > > > In file modules/dm_tests.pm, function trans_unpack, line 782, changed : > > > > my @vars = unpack $type, $packed; > > > > to > > > > my @vars = unpack "$type*", $packed; > > > > (and the same at line 795) > > Hmm, it is working here, it is used in the compaq-server/log/transforms > template file, and I have a number of servers using this template, and the > values work correctly. Can you provide more detail on when/where/how it isn't > working? I am building physical address (MAC) the following way. Don't ask me why I did like this, I don't remember. Full template is attached to this email (see sonic-tz170.tar.bz2). (from transforms) # Build nice Physical address PhysAddrTxtFull : UNPACK : {ifPhysAddr} H PhysAddrTxt1 : SUBSTR : {PhysAddrTxtFull} 0 2 PhysAddrTxt2 : SUBSTR : {PhysAddrTxtFull} 2 2 PhysAddrTxt3 : SUBSTR : {PhysAddrTxtFull} 4 2 PhysAddrTxt4 : SUBSTR : {PhysAddrTxtFull} 6 2 PhysAddrTxt5 : SUBSTR : {PhysAddrTxtFull} 8 2 PhysAddrTxt6 : SUBSTR : {PhysAddrTxtFull} 10 2 (from oids) ifPhysAddr : .1.3.6.1.2.1.2.2.1.6 : branch The output from the device looks like : IF-MIB::ifPhysAddress.1 = STRING: 0:6:b1:1b:12:34 IF-MIB::ifPhysAddress.2 = STRING: IF-MIB::ifPhysAddress.3 = STRING: 0:6:b1:1b:12:35 IF-MIB::ifPhysAddress.4 = STRING: 0:6:b1:1b:12:36 > > 3) SUBSTR did not work > > ... in specific cases. > > (...) > > The fix looks fine. Could you send a diff instead? Sure. See "dm_tests.pm.substr.diff" file attached. Again, thanks for this great piece of software ! Have a great day. -- Gaëtan Frenoy <ga...@fr...> |
From: Kauffman, T. <Kau...@ni...> - 2008-06-09 13:25:16
|
Devmon is hanging for me, on average once per day. No error messages anywhere that I can see, no entries in the log, nothing. All the processes are present. The 'fix' is to kill the master, wait for all the child processes and the master to exit, and restart. I now have a hobbit extension script checking the last dm status and killing devmon if the update was over 4 minutes old. What can I do to get a better handle on the error? Tom Kauffman NIBCO INC CONFIDENTIALITY NOTICE: This email and any attachments are for the exclusive and confidential use of the intended recipient. If you are not the intended recipient, please do not read, distribute or take action in reliance upon this message. If you have received this in error, please notify us immediately by return email and promptly delete this message and its attachments from your computer system. We do not waive attorney-client or work product privilege by the transmission of this message. |
From: Michael A. P. <mp...@sg...> - 2008-06-09 13:35:52
|
A lot of us have this problem, I think its a known-bug. We just setup a crontab to restart once an hour. Let us know if you figure it out, michael On 6/9/08 9:18 AM, "Kauffman, Tom" <Kau...@ni...> wrote: > > > Devmon is hanging for me, on average once per day. No error messages anywhere > that I can see, no entries in the log, nothing. > > All the processes are present. > > The 'fix' is to kill the master, wait for all the child processes and the > master to exit, and restart. I now have a hobbit extension script checking the > last dm status and killing devmon if the update was over 4 minutes old. > > What can I do to get a better handle on the error? > > Tom Kauffman > NIBCO INC > CONFIDENTIALITY NOTICE: This email and any attachments are for the > exclusive and confidential use of the intended recipient. If you are not > the intended recipient, please do not read, distribute or take action in > reliance upon this message. If you have received this in error, please > notify us immediately by return email and promptly delete this message > and its attachments from your computer system. We do not waive > attorney-client or work product privilege by the transmission of this > message. > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Devmon-support mailing list > Dev...@li... > https://lists.sourceforge.net/lists/listinfo/devmon-support > |
From: Alex L. <al...@gm...> - 2008-06-09 22:37:26
|
As an altrenative to crontab job you can have hobbit restart devmon service when it hangs ... in hobbit-alerts.cfg ... HOST=xxx.yyy.com COLOR=purple SERVICE=dm SCRIPT /home/hobbit/bin/devmon_restart.sh 1234567890 ... #cat /home/hobbit/bin/devmon_restart.sh ps -ax | grep devm | grep perl|awk '{print $1}' | xargs kill sleep 60 ps -ax | grep devm | grep perl|awk '{print $1}' | xargs kill sleep 10 /home/hobbit/devmon/devmon On Mon, Jun 9, 2008 at 6:35 AM, Michael A. Price <mp...@sg...> wrote: > A lot of us have this problem, I think its a known-bug. > > We just setup a crontab to restart once an hour. > > Let us know if you figure it out, michael > > > On 6/9/08 9:18 AM, "Kauffman, Tom" <Kau...@ni...> wrote: > > > > Devmon is hanging for me, on average once per day. No error messages > anywhere that I can see, no entries in the log, nothing. > > All the processes are present. > > The 'fix' is to kill the master, wait for all the child processes and the > master to exit, and restart. I now have a hobbit extension script checking > the last dm status and killing devmon if the update was over 4 minutes old. > > What can I do to get a better handle on the error? > > Tom Kauffman > NIBCO INC > CONFIDENTIALITY NOTICE: This email and any attachments are for the > exclusive and confidential use of the intended recipient. If you are not > the intended recipient, please do not read, distribute or take action in > reliance upon this message. If you have received this in error, please > notify us immediately by return email and promptly delete this message > and its attachments from your computer system. We do not waive > attorney-client or work product privilege by the transmission of this > message. > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Devmon-support mailing list > Dev...@li... > https://lists.sourceforge.net/lists/listinfo/devmon-support > > > > |
From: Buchan M. <bg...@st...> - 2008-06-10 11:01:42
|
On Tuesday 10 June 2008 00:37:15 Alex Levin wrote: [...] > On Mon, Jun 9, 2008 at 6:35 AM, Michael A. Price <mp...@sg...> wrote: > > A lot of us have this problem, I think its a known-bug. Well, there was a similar bug in all 0.3.0 prereleases before 0.3.0. That one was known and was fixed, because I could reproduce it. If anyone who is experiencing this one often (my once-a-month recurrence might take a while to track it down) can provide the debug output when it occurs, then I can try and reproduce and fix it ... So, I would say that people are experiencing this issue, but the cause of the bug is as yet unknown ... > > > > We just setup a crontab to restart once an hour. > > > > Let us know if you figure it out, michael If the approach everyone takes is to automatically restart it ... it's going to take a bit longer to track down. (I've dropped the Hobbit list, please try and avoid cross-posts) Regards, Buchan |
From: Buchan M. <bg...@st...> - 2008-06-09 15:40:03
|
On Monday 09 June 2008 15:18:27 Kauffman, Tom wrote: > Devmon is hanging for me, on average once per day. No error messages > anywhere that I can see, no entries in the log, nothing. > > All the processes are present. > > The 'fix' is to kill the master, wait for all the child processes and the > master to exit, and restart. I now have a hobbit extension script checking > the last dm status and killing devmon if the update was over 4 minutes old. > > What can I do to get a better handle on the error? I've seen this twice in the past month. I haven't had an opportunity to track it down. If you can start devmon in debug mode (say in screen), the last output from it when it dies might be useful. If you can't, I will try and reproduce it. Regards, Buchan |
From: G. F. <ga...@fr...> - 2008-06-30 08:02:59
Attachments:
dm_msg.pm.diff
|
Good day or night, On Monday 09 June 2008 15:18:27 Kauffman, Tom wrote: > Devmon is hanging for me, on average once per day. No error messages > anywhere that I can see, no entries in the log, nothing. Same issue here. Frequency is a couple of times per week. But still very very annoying. Le Mon, 9 Jun 2008 17:39:51 +0200, Buchan Milne écrivait: > I've seen this twice in the past month. I haven't had an opportunity to track > it down. > > If you can start devmon in debug mode (say in screen), the last output from it > when it dies might be useful. For those who want to help troubleshooting this further, here is a little "howto" (for debian-based system): 1) Patch code to actually send message to display server when running in debug mode. By default, if you are running devmon in debug mode, it does not send any report to the display server, just display them on screen. This is fine in testing environement but is not applicable if you want to troubleshoot a problem on your production server. So you may want ot apply attached patch to change this behaviour : patch dm_msg.pm dm_msg.pm.diff 2) Install "screen" aptitude install screen 3) Stop any running devmin instance /etc/init.d/devmon stop 4) Running devmon in verbose debug "screened" foreground mode : screen /usr/local/devmon/devmon -f --debug -vvvv 5) Go get some coffee You can close the console devmon is running into and wait for the problem to happen. You can also leave the console running a "detach" from the screen by hitting "Ctrl+A d". 6) Go back to the console To re-attach to the console devmon is running into (certainly when devmon goes purple) : screen -r And send the information you see on this list. As I just set this up on my production machine which is hanging every couple of days, I hope I can send you more information pretty soon. Have a great day or night. -- Gaëtan Frenoy <ga...@fr...> |
From: xbgmsharp <xbg...@gm...> - 2008-06-12 11:48:54
|
Hello, I did get the same problem, plus one :) I upgraded devmon from v0.3 beta4 to v0.3 final About # This is also used by the bb-hostsvc.cgi script to determine if the detailed # status view of a test should include a graph. TEST2RRD need to be set like this TEST2RRD="if_load=devmon" But if you do that you don't get the graph in the status page only in the trends pages. TEST2RRD="if_load" You will get it in the trends and status but it doesn't graph anymore. As anyone get the same problem? Regards >> >> Recently, I upgraded devmon from v0.2.2 to v0.3 final >> (devmon-0.3.0.tar.gz / 2008-04-03 06:52 on Sourceforce). >> >> Below are some hacks I had to do to make it working for me. >> Any comment is of course very welcome. >> >> >> 1) CONN test must be green >> >> All our firewalls have ICMP ping disabled. In our bb-hosts, they are >> all marked as "noping". So the "conn" status of all those firewalls is >> "clear". >> >> With v0.3, it looks that devmon is skipping all devices that are not >> marked as conn=green. > > You can disable this feature globally by changing BBTYPE in devmon.cfg to > something other than "hobbit". I see now this change was not documented in > the CHANGELOG, but it was implemented before I took over maintainership. I don't see the point of changing this parameter as it doesn't about hte documentation. # BBTYPE: Determines how devmon will format its output pages.Accetable values # are 'bb' or 'hobbit', either each option setting the output format to that # of its respective server type [DEFAULT: hobbit] > > At present there is no way to disable it on a per-device basis. > >> This is not applicable for us. I guess we should have a global >> variable that gives all acceptable status (with default = green) but so >> far, I just changed the tests within dm_snmp.pm and dm_tests.pm to >> accept "clear" status. >> >> >> 2) UNPACK did not work >> >> In file modules/dm_tests.pm, function trans_unpack, line 782, changed : >> >> my @vars = unpack $type, $packed; >> >> to >> >> my @vars = unpack "$type*", $packed; >> >> (and the same at line 795) > > Hmm, it is working here, it is used in the compaq-server/log/transforms > template file, and I have a number of servers using this template, and the > values work correctly. Can you provide more detail on when/where/how it isn't > working? > >> 3) SUBSTR did not work > > ... in specific cases. > >> (btw, I have the feeling I already fixed this before...) >> >> In file modules/dm_tests.pm, function trans_substr, line 548 and >> following, changed: >> >> if(defined $length) { >> $oid_h->{'val'}{$leaf} = substr $string, $offset, $length; >> } >> else { >> $oid_h->{'val'}{$leaf} = substr $string, $offset; >> } >> >> to >> >> my $string_length = length($string); >> if ($offset <= $string_length) { >> if(defined $length) { >> if ($offset + $length <= $string_length) { >> $oid_h->{'val'}{$leaf} = substr $string, $offset, >> $length; } else { >> # Length outside string limit, review length >> $oid_h->{'val'}{$leaf} = substr $string, $offset, >> $string_length - $offset; } >> } else { >> $oid_h->{'val'}{$leaf} = substr $string, $offset; >> } >> } else { >> # Offset outside string limit, returns an empty string >> $oid_h->{'val'}{$leaf} = ''; >> } >> >> > > The fix looks fine. Could you send a diff instead? > > >> 4) VERSION is still "rc1" >> > > Yes, I noticed it a few weeks ago. > >> And a last very minor change : >> >> In modules/dm_config.pm, line 38 : changed version from 0.3.0-rc1 to >> 0.3.0 > > I think what I will do is try and fix all the real bugs (no features) in the > next week or three, and release a 0.3.1. > > Thanks for the feedback. > > Regards, > Buchan > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Devmon-support mailing list > Dev...@li... > https://lists.sourceforge.net/lists/listinfo/devmon-support > > -- Thanks for using xbgm# / Devmon / BBwin. http://xbgm.sourceforge.net/ http://devmon.sourceforge.net/ http://bbwin.sourceforge.net/ Please feedback. |
From: Hezki E. <me...@gm...> - 2008-06-12 11:54:27
|
Hi I also had this problem , after some testing I managed to fix that by using : TESTRRD="...if_load=devmon" GRAPHS="...,if_load::1,devmon::1" On Thu, Jun 12, 2008 at 2:48 PM, xbgmsharp <xbg...@gm...> wrote: > Hello, > > I did get the same problem, plus one :) > > I upgraded devmon from v0.3 beta4 to v0.3 final > > About > # This is also used by the bb-hostsvc.cgi script to determine if the > detailed > # status view of a test should include a graph. > TEST2RRD need to be set like this TEST2RRD="if_load=devmon" > > But if you do that you don't get the graph in the status page only in > the trends pages. > TEST2RRD="if_load" > > You will get it in the trends and status but it doesn't graph anymore. > > As anyone get the same problem? > > Regards > > > > >> > >> Recently, I upgraded devmon from v0.2.2 to v0.3 final > >> (devmon-0.3.0.tar.gz / 2008-04-03 06:52 on Sourceforce). > >> > >> Below are some hacks I had to do to make it working for me. > >> Any comment is of course very welcome. > >> > >> > >> 1) CONN test must be green > >> > >> All our firewalls have ICMP ping disabled. In our bb-hosts, they are > >> all marked as "noping". So the "conn" status of all those firewalls is > >> "clear". > >> > >> With v0.3, it looks that devmon is skipping all devices that are not > >> marked as conn=green. > > > > You can disable this feature globally by changing BBTYPE in devmon.cfg to > > something other than "hobbit". I see now this change was not documented > in > > the CHANGELOG, but it was implemented before I took over maintainership. > > > I don't see the point of changing this parameter as it doesn't about > hte documentation. > # BBTYPE: Determines how devmon will format its output pages.Accetable > values > # are 'bb' or 'hobbit', either each option setting the output format to > that > # of its respective server type [DEFAULT: hobbit] > > > > > At present there is no way to disable it on a per-device basis. > > > >> This is not applicable for us. I guess we should have a global > >> variable that gives all acceptable status (with default = green) but so > >> far, I just changed the tests within dm_snmp.pm and dm_tests.pm to > >> accept "clear" status. > >> > >> > >> 2) UNPACK did not work > >> > >> In file modules/dm_tests.pm, function trans_unpack, line 782, changed : > >> > >> my @vars = unpack $type, $packed; > >> > >> to > >> > >> my @vars = unpack "$type*", $packed; > >> > >> (and the same at line 795) > > > > Hmm, it is working here, it is used in the compaq-server/log/transforms > > template file, and I have a number of servers using this template, and > the > > values work correctly. Can you provide more detail on when/where/how it > isn't > > working? > > > >> 3) SUBSTR did not work > > > > ... in specific cases. > > > >> (btw, I have the feeling I already fixed this before...) > >> > >> In file modules/dm_tests.pm, function trans_substr, line 548 and > >> following, changed: > >> > >> if(defined $length) { > >> $oid_h->{'val'}{$leaf} = substr $string, $offset, $length; > >> } > >> else { > >> $oid_h->{'val'}{$leaf} = substr $string, $offset; > >> } > >> > >> to > >> > >> my $string_length = length($string); > >> if ($offset <= $string_length) { > >> if(defined $length) { > >> if ($offset + $length <= $string_length) { > >> $oid_h->{'val'}{$leaf} = substr $string, $offset, > >> $length; } else { > >> # Length outside string limit, review length > >> $oid_h->{'val'}{$leaf} = substr $string, $offset, > >> $string_length - $offset; } > >> } else { > >> $oid_h->{'val'}{$leaf} = substr $string, $offset; > >> } > >> } else { > >> # Offset outside string limit, returns an empty string > >> $oid_h->{'val'}{$leaf} = ''; > >> } > >> > >> > > > > The fix looks fine. Could you send a diff instead? > > > > > >> 4) VERSION is still "rc1" > >> > > > > Yes, I noticed it a few weeks ago. > > > >> And a last very minor change : > >> > >> In modules/dm_config.pm, line 38 : changed version from 0.3.0-rc1 to > >> 0.3.0 > > > > I think what I will do is try and fix all the real bugs (no features) in > the > > next week or three, and release a 0.3.1. > > > > Thanks for the feedback. > > > > Regards, > > Buchan > > > > ------------------------------------------------------------------------- > > Check out the new SourceForge.net Marketplace. > > It's the best place to buy or sell services for > > just about anything Open Source. > > http://sourceforge.net/services/buy/index.php > > _______________________________________________ > > Devmon-support mailing list > > Dev...@li... > > https://lists.sourceforge.net/lists/listinfo/devmon-support > > > > > > -- > Thanks for using xbgm# / Devmon / BBwin. > http://xbgm.sourceforge.net/ > http://devmon.sourceforge.net/ > http://bbwin.sourceforge.net/ > Please feedback. > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Devmon-support mailing list > Dev...@li... > https://lists.sourceforge.net/lists/listinfo/devmon-support > |