July 2012 - syslog-ng - lists.balabit.hu

non-UTF-8 data in mongodb
by Lucas, Sascha 25 Jul '12

25 Jul '12

Hi, I was looking for a way to have some log-rotate functionality (move logs older then x into another collection). I heard of TTL collections, which is fine for deletion. To keep collections small, I tried following mongo-shell code: for( var c = db.messages.find({DATE: {$lt: "1342688698"}}); c.hasNext(); ) { db.messages.old.save(c.next()); } Being totally new to mongodb, I don’t know if this is the right solution, however I got the following error: Mon Jul 23 12:30:34 decode failed. probably invalid utf-8 string [opx2 opx2-dispatch.exe(8100) 2 OPX-20001 Attempt to do an array operation on #26\META-SUPER-HYPER-� which is not an array.] Mon Jul 23 12:30:34 why: TypeError: malformed UTF-8 character sequence at offset 98 Mon Jul 23 12:30:34 Error: invalid utf8 shell/collection.js:128 Asking on mongodb-user Tad Marshall explained that running mongodb with --objcheck prevents this problem. The mogodb config file has following comment: # Inspect all client data for validity on receipt (useful for # developing drivers) #objcheck = true Is the syslog-ng shipped driver the root cause? Should syslog-ng utf8-encode the strings? Is this a bug, or do I miss something? TIA, Sascha. Vorsitzender des Aufsichtsrates: Ralf Hiltenkamp Geschäftsführung: Michael Krüger (Sprecher), Stephan Drescher Sitz der Gesellschaft: Halle/Saale Registergericht: Amtsgericht Stendal | Handelsregister-Nr. HRB 208414 UST-ID-Nr. DE 158253683 Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Empfänger sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail oder des Inhalts dieser Mail sind nicht gestattet. Diese Kommunikation per E-Mail ist nicht gegen den Zugriff durch Dritte geschützt. Die GISA GmbH haftet ausdrücklich nicht für den Inhalt und die Vollständigkeit von E-Mails und den gegebenenfalls daraus entstehenden Schaden. Sollte trotz der bestehenden Viren-Schutzprogramme durch diese E-Mail ein Virus in Ihr System gelangen, so haftet die GISA GmbH - soweit gesetzlich zulässig - nicht für die hieraus entstehenden Schäden.

2 2

Glob character ranges
by T. A. Smooth 24 Jul '12

24 Jul '12

Any one have an example of a glob character range for matching a value? match("*[0-9]" value ("MSG") type ("glob") ) The above should match Dog 1 Dog 2 Dog 3

5 7

Syslog-ng MongoDB inserting date data type
by Michael Dillon 24 Jul '12

24 Jul '12

I'm using Epilog (Snare) to push syslog messages from my IIS logs to syslog-ng. I'm using the CSV parser to parse the message body and create nv pairs. I can't seem to figure out how to get the messages inserted with a datetime data type. Any insights? -- Michael Dillon

2 1

prevent duplicate messages in logs
by Sergey Naumov 23 Jul '12

23 Jul '12

Hello. I would like to ask what mechanism syslog-ng provides to prevent duplicate messages in a log. In most common scenarios I can use "final" flag, for example, when I want only logging on a local machine (so when I have a single destination). But imagine, that user can specify any combination of facilities/priorities and destinations, for example, user want to log auth/>=info and all/>=notice to local machine and auth/>=debug to remote. Then I have a problem: I can't use "final" flag for these entries, because some messages (say auth/>=info) will not be logged to remote host, but only auth/debug, and if I don't use final flag, there will be duplicates of all auth/>=notice messages in the local log. Is there a way to solve this problem in general? One way is to make a separate log entry for each facility/priority combination and group all destinations there, but it won't work if user specify some additional custom filters. Thanks in advance, Sergey Naumov.

2 2

Question regarding lib/nvtable.c (FAIL: test_nvtable)
by John Skopis 20 Jul '12

20 Jul '12

Hello, I just tried to rebuild the EPEL6 SRPM for syslog-ng. The build fails on make check because the test_nvtable test fails. Forgive me if the solution has already been posted. I was able to find a post, but it linked to a few RH bugs suggesting improper use of g_hash_table_insert in nv_registry_alloc_handle(nvtable.c). I looked at nvtable.c and noticed the hash is created: g_hash_table_new_full(g_str_hash, g_str_equal, NULL, NULL) Then, in nv_registry_free: for (i = 0; i < self->names->len; i++) g_free(g_array_index(self->names, NVHandleDesc, i).name); g_array_free(self->names, TRUE); I am wondering if there is a good reason not to pass in g_free as the "key_destroy_func" to g_hash_table_new_full (http://developer.gnome.org/glib/2.29/glib-Hash-Tables.html#g-hash-table-new…) ? Thanks, John

1 0

what's next for syslog-ng?
by Peter Czanik 20 Jul '12

20 Jul '12

Hello, It was in last weeks syslog-ng newsletter, but as I'd like to see more input, I post it again: Sources for some previously Premium Edition-only features have been available for a while now, but have not been merged into the Open Source Edition yet. As we want to release OSE 3.4 soon, we are unlikely to have the necessary resources to merge all of these. We would like to ask Your opinion on which of these features would you like to see in the upcoming release. Please vote and share your reasoning in the BalaBit community forum at http://communities.balabit.com/balabit/topics/poll_which_of_these_features_… Thanks for your feedback in these hot summer days! :) Bye, -- Peter Czanik (CzP) <czanik(a)balabit.hu> BalaBit IT Security / syslog-ng upstream http://czanik.blogs.balabit.com/

2 2

mongodb with auth in OSE 3.3/3.4
by Lucas, Sascha 20 Jul '12

20 Jul '12

Hi, I would like to have mongodb support with auth enabled. With syslog-ng-OSE 3.3.5 this seems not supported? destination d_mongo { mongodb( username("XXXXXXX") password("XXXXXXXXXXXX")); }; results in not connecting right to mongodb (despite no configuration error). As an excuse, not being a dev, I grep-ed these parameters from the source, but not from the docs :). ATM I have no luck with 3.4.0alpha3 and mongodb (even without auth). But what I'm interested in, is to know if "mongodb with auth" will be supported in the 3.4 release? Or what future plans exists? Thanks Sascha. Vorsitzender des Aufsichtsrates: Ralf Hiltenkamp Geschäftsführung: Michael Krüger (Sprecher), Stephan Drescher Sitz der Gesellschaft: Halle/Saale Registergericht: Amtsgericht Stendal | Handelsregister-Nr. HRB 208414 UST-ID-Nr. DE 158253683 Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Empfänger sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail oder des Inhalts dieser Mail sind nicht gestattet. Diese Kommunikation per E-Mail ist nicht gegen den Zugriff durch Dritte geschützt. Die GISA GmbH haftet ausdrücklich nicht für den Inhalt und die Vollständigkeit von E-Mails und den gegebenenfalls daraus entstehenden Schaden. Sollte trotz der bestehenden Viren-Schutzprogramme durch diese E-Mail ein Virus in Ihr System gelangen, so haftet die GISA GmbH - soweit gesetzlich zulässig - nicht für die hieraus entstehenden Schäden.

2 1

reliable log transfer with syslog-ng-OSE (a bit like PEs' diskbuffering)
by Lucas, Sascha 19 Jul '12

19 Jul '12

Hi, I'm trying to have reliable log transfer (in the sense of network failure) with syslog-ng-OSE. Here my idea (syslog-ng.conf): ## version='3.3.5' ## first save every message to a file source s_relay { tcp( max-connections(10) port(4321) ); udp( port(4321) ); }; destination d_buffer { file("/data/syslog-buffer" template("$ISODATE $HOST $SOURCEIP $PROGRAM $PID $PRI $MSG\n") ); }; log { source(s_relay); destination(d_buffer); }; ## then follow this file and parse it with flow-control enabled source s_buffer { file("/data/syslog-buffer" flags(no-parse) follow_freq(0.1)); }; parser p_buffer { csv-parser( columns("buffer.isodate", "buffer.host", "buffer.srcip", "buffer.program", "buffer.pid", "buffer.pri", "buffer.msg") delimiters(" ") flags(greedy) ); }; destination d_mongo_unrel { mongodb(...); }; log { source(s_buffer); parser(p_buffer); destination(d_mongo_unrel); flags(flow-control); }; What do you think about this solution? Are there objections? I need to add follow_freq() in s_buffer. Without it, the mongodb fills a bit slow. But I don't understand why. Can someone give me a hint, how to get the actual reading position from s_buffer (i.e. from syslog-ng.persist?) and then to truncate it (shell commands preferred)? TIA, Sascha. Vorsitzender des Aufsichtsrates: Ralf Hiltenkamp Geschäftsführung: Michael Krüger (Sprecher), Stephan Drescher Sitz der Gesellschaft: Halle/Saale Registergericht: Amtsgericht Stendal | Handelsregister-Nr. HRB 208414 UST-ID-Nr. DE 158253683 Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Empfänger sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail oder des Inhalts dieser Mail sind nicht gestattet. Diese Kommunikation per E-Mail ist nicht gegen den Zugriff durch Dritte geschützt. Die GISA GmbH haftet ausdrücklich nicht für den Inhalt und die Vollständigkeit von E-Mails und den gegebenenfalls daraus entstehenden Schaden. Sollte trotz der bestehenden Viren-Schutzprogramme durch diese E-Mail ein Virus in Ihr System gelangen, so haftet die GISA GmbH - soweit gesetzlich zulässig - nicht für die hieraus entstehenden Schäden.

2 1

Draft and incomplete documentation for syslog-ng OSE 3.4
by Fekete Robert 19 Jul '12

19 Jul '12

Hi Everyone, I have started to update the Administrator Guide for version 3.4. It is far from finished, and several things are still missing, but as I won't be able to work on it in August, I have released it as it is, since it already contains some updates and useful information. You can find it at http://www.balabit.com/sites/default/files/documents/syslog-ng-ose-3.4-guid… The list of new stuff that is already documented is available at: http://www.balabit.com/sites/default/files/documents/syslog-ng-ose-3.4-guid… Some other smaller changes and corrections is listed here: http://www.balabit.com/sites/default/files/documents/syslog-ng-ose-3.4-guid… I hope to get back to finishing the updates early September, but this depends on other things as well. Regards, Robert

1 0

Fwd: Re: Following a file to generate one syslog message per appended line
by Terry Burton 18 Jul '12

18 Jul '12

---------- Forwarded message ---------- From: "Terry Burton" <tez(a)terryburton.co.uk> Date: 17 Jul 2012 19:52 Subject: Re: [syslog-ng] Following a file to generate one syslog message per appended line To: "Balazs Scheidler" <bazsi77(a)gmail.com> On 17 Jul 2012 19:30, "Balazs Scheidler" <bazsi77(a)gmail.com> wrote: > > ----- Original message ----- > > On 17 July 2012 15:20, Jim Hendrick <jrhendri(a)maine.rr.com> wrote: > > > hmmm - nothing obvious to me. > > > > > > Questions that might help > > > - do you know what the message rate for that source is? > > > > The source is the alerts from an IDS where the messages arrive at > > irregular intervals and often in surges of many messages. > > > > > - is there anything possibly unusual about the messages themselves? > > > (how is end of line demarked? what are the maximum line lengths?) > > > > The end of line is demarked by a LF (0x0a). > > > > $ xxd /srv/snort/snort.fast > > > > 0000000: 3037 2f31 362d 3134 3a30 333a 3431 2e33 07/16-14:03:41.3 > > 0000010: 3837 3339 3020 205b 2a2a 5d20 5b31 3a31 87390 [**] [1:1 > > 0000020: 3535 3137 3a31 305d 203c 6574 6831 3e20 5517:10] <eth1> > > 0000030: 5745 422d 434c 4945 4e54 204d 6963 726f WEB-CLIENT Micro > > 0000040: 736f 6674 2057 696e 646f 7773 2041 5649 soft Windows AVI > > 0000050: 2044 6972 6563 7453 686f 7720 5175 6963 DirectShow Quic > > 0000060: 6b54 696d 6520 7061 7273 696e 6720 6f76 kTime parsing ov > > 0000070: 6572 666c 6f77 2061 7474 656d 7074 205b erflow attempt [ > > 0000080: 2a2a 5d20 5b43 6c61 7373 6966 6963 6174 **] [Classificat > > 0000090: 696f 6e3a 2041 7474 656d 7074 6564 2055 ion: Attempted U > > 00000a0: 7365 7220 5072 6976 696c 6567 6520 4761 ser Privilege Ga > > 00000b0: 696e 5d20 5b50 7269 6f72 6974 793a 2031 in] [Priority: 1 > > 00000c0: 5d20 7b54 4350 7d20 322e 3232 2e32 3238 ] {TCP} 2.22.228 > > 00000d0: 2e34 303a 3830 202d 3e20 3134 332e 3231 .40:80 -> 143.21 > > 00000e0: 302e 3936 2e31 3138 3a36 3238 3732 0a30 0.96.118:62872.0 > > 00000f0: 372f 3136 2d31 343a 3033 3a34 312e 3537 7/16-14:03:41.57 > > 0000100: 3836 3535 2020 5b2a 2a5d 205b 313a 3135 8655 [**] [1:15 > > 0000110: 3531 373a 3130 5d20 3c65 7468 313e 2057 517:10] <eth1> W > > 0000120: 4542 2d43 4c49 454e 5420 4d69 6372 6f73 EB-CLIENT Micros > > 0000130: 6f66 7420 5769 6e64 6f77 7320 4156 4920 oft Windows AVI > > 0000140: 4469 7265 6374 5368 6f77 2051 7569 636b DirectShow Quick > > 0000150: 5469 6d65 2070 6172 7369 6e67 206f 7665 Time parsing ove > > 0000160: 7266 6c6f 7720 6174 7465 6d70 7420 5b2a rflow attempt [* > > 0000170: 2a5d 205b 436c 6173 7369 6669 6361 7469 *] [Classificati > > 0000180: 6f6e 3a20 4174 7465 6d70 7465 6420 5573 on: Attempted Us > > 0000190: 6572 2050 7269 7669 6c65 6765 2047 6169 er Privilege Gai > > 00001a0: 6e5d 205b 5072 696f 7269 7479 3a20 315d n] [Priority: 1] > > 00001b0: 207b 5443 507d 2032 2e32 322e 3232 382e {TCP} 2.22.228. > > 00001c0: 3430 3a38 3020 2d3e 2031 3433 2e32 3130 40:80 -> 143.210 > > 00001d0: 2e39 362e 3131 383a 3632 3837 320a 3037 .96.118:62872.07 > > 00001e0: 2f31 362d 3134 3a30 333a 3530 2e39 3534 /16-14:03:50.954 > > 00001f0: 3939 3620 205b 2a2a 5d20 5b31 3a35 3930 996 [**] [1:590 > > 0000200: 333a 395d 203c 6574 6831 3e20 5350 5957 3:9] <eth1> SPYW > > 0000210: 4152 452d 5055 5420 4164 7761 7265 2064 ARE-PUT Adware d > > 0000220: 6f77 6e6c 6f61 6420 6163 6365 6c65 7261 ownload accelera > > 0000230: 746f 7220 706c 7573 2072 756e 7469 6d65 tor plus runtime > > 0000240: 2064 6574 6563 7469 6f6e 202d 2067 6574 detection - get > > 0000250: 2061 6473 205b 2a2a 5d20 5b43 6c61 7373 ads [**] [Class > > 0000260: 6966 6963 6174 696f 6e3a 204d 6973 6320 ification: Misc > > 0000270: 6163 7469 7669 7479 5d20 5b50 7269 6f72 activity] [Prior > > 0000280: 6974 793a 2033 5d20 7b54 4350 7d20 3134 ity: 3] {TCP} 14 > > 0000290: 332e 3231 302e 3139 332e 3639 3a34 3937 3.210.193.69:497 > > 00002a0: 3535 202d 3e20 3231 322e 3134 332e 3232 55 -> 212.143.22 > > 00002b0: 2e31 3130 3a38 300a 3037 2f31 362d 3134 .110:80.07/16-14 > > 00002c0: 3a30 333a 3536 2e34 3536 3034 3320 205b :03:56.456043 [ > > 00002d0: 2a2a 5d20 5b31 3a31 3535 3137 3a31 305d **] [1:15517:10] > > 00002e0: 203c 6574 6831 3e20 5745 422d 434c 4945 <eth1> WEB-CLIE > > 00002f0: 4e54 204d 6963 726f 736f 6674 2057 696e NT Microsoft Win > > 0000300: 646f 7773 2041 5649 2044 6972 6563 7453 dows AVI DirectS > > 0000310: 686f 7720 5175 6963 6b54 696d 6520 7061 how QuickTime pa > > 0000320: 7273 696e 6720 6f76 6572 666c 6f77 2061 rsing overflow a > > 0000330: 7474 656d 7074 205b 2a2a 5d20 5b43 6c61 ttempt [**] [Cla > > 0000340: 7373 6966 6963 6174 696f 6e3a 2041 7474 ssification: Att > > 0000350: 656d 7074 6564 2055 7365 7220 5072 6976 empted User Priv > > 0000360: 696c 6567 6520 4761 696e 5d20 5b50 7269 ilege Gain] [Pri > > 0000370: 6f72 6974 793a 2031 5d20 7b54 4350 7d20 ority: 1] {TCP} > > 0000380: 3233 2e36 372e 3235 352e 3536 3a38 3020 23.67.255.56:80 > > 0000390: 2d3e 2031 3433 2e32 3130 2e31 3030 2e37 -> 143.210.100.7 > > 00003a0: 333a 3535 3539 330a 3037 2f31 362d 3134 3:55593.07/16-14 > > 00003b0: 3a30 343a 3434 2e36 3232 3930 3820 205b :04:44.622908 [ > > 00003c0: 2a2a 5d20 5b31 3a31 3131 3932 3a31 325d **] [1:11192:12] > > <...snip...> > > > > > I would personally try removing the flags (from source and log lines) > > > and one by one add them back - looking at the changes in behavior (if > > > any), then taking that one back out and adding the other. > > > > I've tried that to no useful effect unfortunately. It appears that > > flush_lines(1) is the default setting. > > > > I am wondering whether the file source driver does not treat a lone LF > > as a new line and therefore log_fetch_limit(1) is ineffective? I am > > about to test this. > > it should treat one NL (aka LF, ASCII 10) as line terminator. what version of syslog-ng are you using? Version 3.1.3 on Debian Squeeze, 64 bit.

3 2