knot-dns-users

knot-dns-users@lists.nic.cz

5 participants
727 discussions

by Marek Vavruša

Hi Everyone, we're getting closer to the 1.5 release and because we're mostly feature complete now, I think today might be a good opportunity to share some of the new stuff we've created and hear out your opinions. There's a ton of improvements that I can't cover in a few paragraphs, but I've picked 3 of those that I think are the pillars for the bigger plan for this year. Just so you know, there is no tarball this time, but if you're brave enough, you can actually try everything covered here. You can find the code in our git repository under a tag "v1.5.0-alpha": $ git clone -b v1.5.0-alpha https://gitlab.labs.nic.cz/labs/knot.git 1. Query processing and code base We've known for a while that there were parts of the code that we weren't happy with. It's not just that it was getting hard to patch and read, but with we had a ton of ideas that we couldn't get implemented because of the limitations (in fact, this effort bears fruits as soon as of this release, but more on that later). One of those places was query processing and dealing with queries in general. Naturally, we've sat down and reworked this from scratch. Now, from the user perspective it's not so exciting - yes we made it RFC compliant in several cases and it's faster than ever (I'm running benchmarks as we speak, so you might want to check out our webpage later). It's still a preview and we're ironing things out, so don't take the numbers as final. For those who are interested in the details (thank you!), it's much much more important because the code lays a solid foundation for the cool stuff to come while being lightweight. How much? Around 10+k LOC which is roughly 10-15% of our code base gone and we offset that by adding almost 5k LOC worth of test cases. To put it in another words, it means it does more with less and we're not done here. 2. Separating Knot DNS library We also improved the Knot DNS build process overall. The code providing basic building blocks for the server and utilities is now held in the libknot library, which is by default linked dynamically to these components. You will also notice the libzscanner library, our high-performance zone scanner written in Ragel (with no dependencies on Knot DNS). We plan to build some other projects on top of that and provide public headers in the future (which is more likely to happen for the libzscanner - if someone is interested). But currently, you may benefit from a smaller size of the binaries. 3. Pluggable query processing modules + auto forward/reverse records New features are always a compromise - usually between ease of use, complexity and resource usage. But what if we could make the core as lightweight as possible and let the user choose? The new processing API is built like a Lego (don't step on it though), it breaks down the processing into independent parts that together form a plan for query processing. So what we did was, we've created an interface for loadable modules and let it add/remove or change the query processing plan as the module sees fit. This allows to keep complexity at bay (for those who don't want extra stuff) and provide a playground for new ideas. We actually have a first module - it can synthesize per-answer forward/reverse records for IPv4/IPv6 addresses from a template in configuration file (see 4.11 in the documentation on further information). For the impatient, an example speaks the best: We have an IPv6 reverse zone we need to take care of, but the problem is the address space is too big to generate a zone file. So we create an empty zone, put in static records we want and let the module to synthesize other records per-answer (if they fit in the template and aren't present in the zone): # Configuration: 1.6.b.0.0.0.0.0.0.2.6.2.ip6.arpa { query_module { synth_record "reverse dynamic- example. 400 2620:0:b61::/52"; } Now if we load Knot DNS, and kdig for a reverse record, we get a following reply: $ kdig PTR 1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.1.6.b.0.0.0.0.0.0.2.6.2.ip6.arpa. ;; QUESTION SECTION: ;; 1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.1.6.b.0.0.0.0.0.0.2.6.2.ip6.arpa. 0 IN PTR ;; ANSWER SECTION: ..snip.. 400 IN PTR dynamic-2620-0000-0b61-0000-0000-0000-0000-0001.example. Forward records can be dealt with in a similar fashion, see the manual 4.11 for more examples and information. To sum it up, modules could be a way to accommodate for the "should this be a in a nameserver?" sort of things. If you don't like it, just don't put it in the configuration and you're good. As of now the module API is open (no "how to" though), and so are we for ideas. Phew, this was longer than I expected and I barely scratched the surface. As you see, this version is going to be a lot about tidying up and opening up the possibilities. So what's your take on this? Best, Marek -- Marek Vavrusa, Knot DNS CZ.NIC Labs http://www.knot-dns.cz ------------------------------------------- Americká 23, 120 00 Praha 2, Czech Republic WWW: http://labs.nic.cz http://www.nic.cz

11 let

Knot DNS 1.4.4

by Marek Vavruša

Hi Everyone, while we in the CZ.NIC Labs are all focused on the next big thing, I thought now it might be a good time to get out another of our (admittedly highly irregular) monthly patch releases. Bug fixes aside, there are a couple improvements that we hope will make your life easier. Here's the breakdown: - 'knotc reload' does not immediately refresh/notify unchanged zones. This fixes the flurry of refresh messages after each reload and makes it lightweight enough to be used frequently. Note that new/changed zones are still added and old zones are removed. However, if you want to refresh zones explicitly, you need to call 'knotc refresh' after reload now. - 'knotc -f refresh' now truly forces a zone refresh. To put it in another words, it's akin to the 'retransfer' command that you missed in the Knot DNS and instead of checking for new zone, it starts a full transfer of the zone right away. - 'knotc' remote commands are now logged in the daemon logfile Now, there are several bugfixes as well. See the NEWS for a complete list, but here's a selection of the most notable - in several cases notify messages weren't sent after a zone resign, progressive bootstrap retry regression was fixed, few issues with journal and maximum entry size. There is also a slight behaviour change in the zone file parser and the daemon itself. First - zone file parser now accepts asterisk in the domain name labels (wildcards aside). As for the daemon, if a zone is in a slave mode and fails to load for some reason, it immediately tries to reboostrap from the master server instead of just reporting an error. I'd also like to thank to Robert S. Edmonds for amending various spelling errors and typos in manpages and documentation, thanks! So that's it, I hope the improvements make this update a little worthwhile before new stuff comes out later this spring. Full changelog: https://gitlab.labs.nic.cz/labs/knot/blob/v1.4.4/NEWS Sources: https://secure.nic.cz/files/knot-dns/knot-1.4.4.tar.gz https://secure.nic.cz/files/knot-dns/knot-1.4.4.tar.xz GPG signatures: https://secure.nic.cz/files/knot-dns/knot-1.4.4.tar.gz.asc https://secure.nic.cz/files/knot-dns/knot-1.4.4.tar.xz.asc Kind Regards, Marek -- Marek Vavrusa, Knot DNS CZ.NIC Labs http://www.knot-dns.cz ------------------------------------------- Americká 23, 120 00 Praha 2, Czech Republic WWW: http://labs.nic.cz http://www.nic.cz

11 let

Some issues with dnssec signed domain.

by Josef Karliak

Hi there, I migrated our primary DNS from Bind to Knot. I runned some tests by nic.cz's dnscheck, but there is an error: DNSSEC signature RRSIG(fnhk.cz/IN/SOA/64431) fails to validate the RR set: key 1: keytag does not match key 2:RSA Verification failed Link to test: http://dnscheck.labs.nic.cz/?time=1395821962&id=102810&view=advanced&test=s… Knot doesn't complains to anything in the system log, fnhk.cz zone is succefully signed. What did I missed ? Thanks and best regards J.Karliak.

11 let

[PATCH] various spelling and typo fixes

by Robert S. Edmonds

--- Doxy.page.h | 2 +- doc/introduction.texi | 2 +- man/kdig.1.in | 2 +- man/khost.1.in | 2 +- src/utils/common/params.h | 2 +- src/utils/dig/dig_exec.c | 4 ++-- src/utils/dig/dig_params.h | 4 ++-- src/utils/nsupdate/nsupdate_params.h | 2 +- 8 files changed, 10 insertions(+), 10 deletions(-) diff --git a/Doxy.page.h b/Doxy.page.h index 03e43e5..e1b91f2 100644 --- a/Doxy.page.h +++ b/Doxy.page.h @@ -33,7 +33,7 @@ Knot DNS supports the following DNS features: - AXFR - master, slave - IXFR - master (primary master experimental), slave - TSIG -- ENDS0 +- EDNS0 - DNSSEC, including NSEC3 - NSID - Unknown RR types diff --git a/doc/introduction.texi b/doc/introduction.texi index fccb2f1..96e53cb 100644 --- a/doc/introduction.texi +++ b/doc/introduction.texi @@ -31,7 +31,7 @@ Knot DNS supports the following DNS features: @item TCP/UDP protocols @item AXFR, IXFR - master, slave @item TSIG -@item ENDS0 +@item EDNS0 @item DNSSEC, including NSEC3 @item NSID @item Unknown RR types diff --git a/man/kdig.1.in b/man/kdig.1.in index 69a19e8..c313411 100644 --- a/man/kdig.1.in +++ b/man/kdig.1.in @@ -151,7 +151,7 @@ Show TTL value. Use TCP protocol (default is UDP for standard query and TCP for AXFR/IXFR). .TP .BR + [ no ] fail -Stop quering next nameserver if SERVFAIL response is received. +Stop querying next nameserver if SERVFAIL response is received. .TP .BR + [ no ] ignore Don't use TCP automatically if truncated reply is received. diff --git a/man/khost.1.in b/man/khost.1.in index e25ba35..1bf677b 100644 --- a/man/khost.1.in +++ b/man/khost.1.in @@ -41,7 +41,7 @@ Print help. Disable recursion. .TP .B \-s -Stop quering next nameserver if SERVFAIL response is received. +Stop querying next nameserver if SERVFAIL response is received. .TP .B \-T Use TCP protocol. diff --git a/src/utils/common/params.h b/src/utils/common/params.h index 78dd70a..d7c34c6 100644 --- a/src/utils/common/params.h +++ b/src/utils/common/params.h @@ -98,7 +98,7 @@ typedef struct { bool hide_cname; } style_t; -/*! \brief Parametr handler. */ +/*! \brief Parameter handler. */ typedef int (*param_handle_f)(const char *arg, void *params); /*! \brief Parameter argument type. */ diff --git a/src/utils/dig/dig_exec.c b/src/utils/dig/dig_exec.c index aaab062..e10b7a8 100644 --- a/src/utils/dig/dig_exec.c +++ b/src/utils/dig/dig_exec.c @@ -409,7 +409,7 @@ static void process_query(const query_t *query) WALK_LIST(server, query->servers) { srv_info_t *remote = (srv_info_t *)server; - DBG("Quering for owner(%s), class(%u), type(%u), server(%s), " + DBG("Querying for owner(%s), class(%u), type(%u), server(%s), " "port(%s), protocol(%s)\n", query->owner, query->class_num, query->type_num, remote->name, remote->service, get_sockname(socktype)); @@ -652,7 +652,7 @@ static void process_query_xfr(const query_t *query) // Use the first nameserver from the list. srv_info_t *remote = HEAD(query->servers); - DBG("Quering for owner(%s), class(%u), type(%u), server(%s), " + DBG("Querying for owner(%s), class(%u), type(%u), server(%s), " "port(%s), protocol(%s)\n", query->owner, query->class_num, query->type_num, remote->name, remote->service, get_sockname(socktype)); diff --git a/src/utils/dig/dig_params.h b/src/utils/dig/dig_params.h index 7287fb7..c9caa3b 100644 --- a/src/utils/dig/dig_params.h +++ b/src/utils/dig/dig_params.h @@ -90,7 +90,7 @@ typedef struct { int32_t wait; /*!< Ignore truncated response. */ bool ignore_tc; - /*!< Stop quering if servfail. */ + /*!< Stop querying if servfail. */ bool servfail_stop; /*!< Class number (16unsigned + -1 uninitialized). */ int32_t class_num; @@ -116,7 +116,7 @@ typedef struct { /*! \brief Settings for dig. */ typedef struct { - /*!< Stop processing - just pring help, version,... */ + /*!< Stop processing - just print help, version,... */ bool stop; /*!< List of DNS queries to process. */ list_t queries; diff --git a/src/utils/nsupdate/nsupdate_params.h b/src/utils/nsupdate/nsupdate_params.h index 599c9b4..e823f75 100644 --- a/src/utils/nsupdate/nsupdate_params.h +++ b/src/utils/nsupdate/nsupdate_params.h @@ -43,7 +43,7 @@ /*! \brief nsupdate-specific params data. */ typedef struct { - /*!< Stop processing - just pring help, version,... */ + /*!< Stop processing - just print help, version,... */ bool stop; /*!< List of files with query data. */ list_t qfiles; -- 1.9.0

11 let

Audit the consistency of zone files, out of sync MX

by Maren S. Leizaola

Hello, What do you guys recommend to audit every resource record in a zone file against all the records in all the DNS. I want something that I feed the master zone file and then goes to each NS server and ensures that the records are correct in all of them. For some strange reason all my DNS servers have the same SOA Serial, but after deleting two MX records, some 4 out of 5 the DNS servers have not taken this update. I've yet to figure out the cause, but I see that SOA Serial is not to be trusted. Regards, Maren.

11 let, 1 měsíc

Re: [knot-dns-users] knot-dns-users Digest, Vol 28, Issue 9

by Maren S. Leizaola

On 2/24/2014 7:00 PM, knot-dns-users-request(a)lists.nic.cz wrote: > Date: Mon, 24 Feb 2014 09:16:27 +0100 > From: Jan Kadlec <jan.kadlec(a)nic.cz> > To: "Maren S. Leizaola" <leizaola(a)hk.com> > Cc: knot-dns-users(a)lists.nic.cz > Subject: Re: [knot-dns-users] GLUE and additional records. > Message-ID: > <1393229787.1717.12.camel(a)labs.jan.kadlec.ws.eth.1.office.nic.cz> > Content-Type: text/plain; charset="UTF-8" > > Hello Maren and thanks for your report. Knot normally sends glue records > in additional section, it seems as if you might have encountered a bug. > Could you provide more data? One NS, A (AAAA) combination and a dig > output for this combination should be enough. Thanks again, Jan. > The environment is as follows. I host hk.com on 6 DNS servers, right now 5 are bind and one is Knot. hk.com's name servers are a,b,c,d,e,f.udrtld.net b is running Knot. Try this link www.intodns.com/hk.com This reports that B provides no glue. Please ignore the errors on f. i've yet not setup urdtld.net on it. dig -cA hk.com @b.udrtld.net when I do a dig -cA hk.com @a.udrtld.net Am I making any mistakes here? Maren.

11 let, 1 měsíc

GLUE and additional records.

by Maren S. Leizaola

Hello, I am currently testing Knot on our zones and find that it does not give any of the additional records which contain the IP of the names servers ie knot servers donot provide Glue records. This is one good thing that Bind does at it reduces the number of queries a resolver has to make. Is there any way for us to be able to do this. Regards, Maren.

11 let, 1 měsíc

/help

by Maren S. Leizaola

/subscribe /list

11 let, 1 měsíc

Knot DNS 1.4.3

by Marek Vavruša

Hi Everyone, I think the time is ripe for a small status update, preview of things to come and a new patch release. Sounds good? So, at the moment, our team at the CZ.NIC Labs is working towards the next major thing, but I'd like to reiterate that we didn't forget about the 1.4 and as of today we have a new release with several important backported bugfixes, here's the brief version: * Two bugs related to authenticated denial of existence proof with a certain combination of wildcard expansions triggering an assertion failure * Comparison of $ORIGIN and zone file in configuration is case insensitive * Corrupted journal data caused a cleanup failure during the zone loading In addition to the bugfixes, we have also slipped in a small enhancement - "include" statement in the configuration can include whole directory, this is useful if you have really a lot of includes in one place. Now a brief status update on what are we working on and what can you expect in the next few weeks/months. The next major release is going to happen later this spring and it's going to be focused on three things mostly - covering the code by tests, cleaning up and polishing existing features. There's going to be a feature or two, but that is not the main focus. The reason is the functionality we have piled up over the time has taken it's toll and now is the right time to sit back and make the features leaner and meaner (that includes usability as well), as that was always an integral idea behind Knot DNS. Plus less things means less things to break, right? I'd like to thank namely Petr Stastny and Olafur Gudmundsson (among others) for bugreports, ideas and thoughts. Full changelog: https://gitlab.labs.nic.cz/labs/knot/blob/v1.4.3/NEWS Sources: https://secure.nic.cz/files/knot-dns/knot-1.4.3.tar.gz https://secure.nic.cz/files/knot-dns/knot-1.4.3.tar.xz GPG signatures: https://secure.nic.cz/files/knot-dns/knot-1.4.3.tar.gz.asc https://secure.nic.cz/files/knot-dns/knot-1.4.3.tar.xz.asc Kind Regards, Marek -- Marek Vavrusa, Knot DNS CZ.NIC Labs http://www.knot-dns.cz ------------------------------------------- Americká 23, 120 00 Praha 2, Czech Republic WWW: http://labs.nic.cz http://www.nic.cz

11 let, 2 měsíce

Feature request: force zone retransfer

by Anand Buddhdev

Hi Knot developers, I have a feature request: I'd like a knotc command that will force Knot to transfer a zone that is configured as a slave. The only command I can issue at the moment is "refresh", but this will not transfer a zone if Knot has a higher serial than the master. Such a forced transfer command is useful to help recover from situations where due to some operational error on the master server, the serial number has gone back, and we deliberately need our Knot server to sync to that old serial number. Without this command, the only way to make Knot recover is to stop it, delete the zone file from disk and start it again. But this is not ideal on a running server with lots of zones, because the stop and start times can be quite high (in our case, 37s to stop, and 1m30s to start). Both BIND and NSD provide such commands to ignore the serial number check and force zone transfers. I'd love to see such a command for Knot too. Regards, Anand

11 let, 2 měsíce

← Newer
1
...
58
59
60
61
62
63
64
...
73
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

knot-dns-users