openldap-bugs

openldap-bugs@openldap.org

1 participants
21001 discussions

Re: (ITS#9088) Concurrency not seen in OpenLDAP 2.4.44
by quanah＠symas.com 23 Sep '19

23 Sep '19

--On Monday, September 23, 2019 10:22 AM +0000 jrajalakshmi(a)juniper.net wrote: > Full_Name: Rajalakshmi > Version: 2.4.44 > OS: RHEL 7.6 > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (116.197.184.12) Hello, RedHat makes changes to OpenLDAP that are not reviewed by the OpenLDAP project. I would advise using an unmodified version of the current release of OpenLDAP to confirm whether or not this issue is in OpenLDAP itself or is caused by a change made by RedHat. You can use the packages from the LTB project for testing: <https://ltb-project.org/documentation/openldap-rpm#yum_repository> Regards, Quanah -- Quanah Gibson-Mount Product Architect Symas Corporation Packaged, certified, and supported LDAP solutions powered by OpenLDAP: <http://www.symas.com>

1 0

Re: (ITS#9055) contrib/slapd-modules/passwd/totp improvements
by ondra＠mistotebe.net 23 Sep '19

23 Sep '19

On Fri, Sep 13, 2019 at 11:57:14AM -0400, Greg Veldman wrote: > On Mon, Sep 09, 2019 at 04:01:59PM +0200, Ond??ej Kuzn??k wrote: >> I mean the ber_str2bv(,,1,) in both new functions. Not sure which code >> you think would overwrite parts of the buffer? ber_str2bv(,,0,) never >> touches it, manually initialising the berval certainly wouldn't either. >> And then you have fewer memory regions to scrub. >> >> Since you already know the length, you can also pass it in so ber_str2bv >> can skip its strlen() check (and since anything can be in a {PLAINTEXT} >> password, you're now embedded NUL safe). > > Ah, OK, I didn't realize that would be NUL safe. I made an > updated patch with that change[1]. > >> I think I mentioned this before as something worth changing: rather >> than call time(0L), you can use op->o_time which is stable and the >> closest you can get to the time the operation was received. > > Yes, sorry I did see that before just forgot to do it. It's > also included in the latest update[1]. Hi Greg, thanks for both, I should merge that soon. On a side note, any ideas how to deal with ppolicy's pwdHistory here so it can reject changing the password to an old one? AFAIK using these schemas will prevent slap_passwd_check() from working and there isn't an obvious way to proceed. Thanks, -- Ondřej Kuzník Senior Software Engineer Symas Corporation http://www.symas.com Packaged, certified, and supported LDAP solutions powered by OpenLDAP

1 0

(ITS#9088) Concurrency not seen in OpenLDAP 2.4.44
by jrajalakshmi＠juniper.net 23 Sep '19

23 Sep '19

Full_Name: Rajalakshmi Version: 2.4.44 OS: RHEL 7.6 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (116.197.184.12) (1) What problem/issue/behavior are you having trouble with? What do you expect to see? Using OpenLDAP version 2.4.44 we are not seeing Concurrency whereas in OpenLDAP Version 2.4.23 we are seeing Concurrent LDAP Requests. In our product OpenLDAP 2.4.23 was previously used and the concurrent request for search using the API ' ldap_search_ext_s() ' was working as expected and our product scalability was better. We have recently upgraded LDAP to OpenLDAP 2.4.44. But with this version, the api 'ldap_search_ext_s()' does not seem to work as expected, i.e concurrent requests handling is not happening. (2) Where are you experiencing the behavior? What environment? In RHEL 7.6. If we sent 10 Concurrent Requests each request is processed one by one. Most of the threads are waiting in ldap_search_req. (3) When does the behavior occur? Frequently? Repeatedly? At certain times? Repeatedly. (4) What information can you provide around timeframes and the business impact? There is a release impact due to the open issue. Hence seeking a reply at the earliest. In our product, the API ldap_search_req is used and this will be called by multiple threads at a time. On the process startup of the application, bind to ldap Server is created. status = ldap_search_ext_s(m_ld, pszBase, nScope, pszFilter, ppszAttrs2, nAttrsOnly, NULL, NULL, pTimeout2, LDAP_NO_LIMIT, ppResult); (5) Whether we need to do anything on that? Request for a solution at the earliest, is there any alternative api to support concurrency request Is there any open defect in the 'OpenLDAP 2.4.44' related to the synchronous handling method ' ldap_search_ext_s() '.

1 0

Re: (ITS#9084) Build issue when slaps is not included
by aixtools＠felt.demon.nl 22 Sep '19

22 Sep '19

This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --gJgclWXWHobp9Jm8TzWLGKLDSFUeHbtMu Content-Type: multipart/mixed; boundary="ftnTy1YEAYlvdoLUzJ64mQJw2X9PqF44q"; protected-headers="v1" From: Michael <aixtools(a)felt.demon.nl> To: openldap-its(a)OpenLDAP.org Cc: aixtools(a)felt.demon.nl Message-ID: <ab713cc5-89fb-8603-6a09-bc95e732c5cf(a)felt.demon.nl> Subject: Re: (ITS#9084) Build issue when slaps is not included --ftnTy1YEAYlvdoLUzJ64mQJw2X9PqF44q Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Content-Language: en-US Update. After uninstall openldap-2.4.46 and openldap-2.4.48 build fine. --ftnTy1YEAYlvdoLUzJ64mQJw2X9PqF44q-- --gJgclWXWHobp9Jm8TzWLGKLDSFUeHbtMu Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- iQEzBAEBCAAdFiEE4MC4YtVFAh73PAYk4vWDeFkjp+AFAl2H0lgACgkQ4vWDeFkj p+CPrQf9Encsrt1sAI2T7VmR5MenhrXDQrZFeiMPjx7aILExkvu68yOSaM195nii yci/MQJWY982I1tA7EFuXKlN0NXM0LKTMK1FpjwrgkITQK6ov2KSPaLV3B9JoQSw MEQyCXhTIiN6ahAUWUw+pVqXk4o313mqqyfmQWoVrUxhgqiZlGFvyStSL+Fp5P+7 7YwxyfvhmEcQsolyqB/WLkw0T86jqPRgk/S8IV69NFiXd3A26bfVyA1sjrlWyRN4 5MRXg/btAimfvH+xlpAebY49GZ1MGt+AdZUu4qr1i2FlJi67Y2Ron/cq5EmwZeYu kqiD1Z5MOd3Ovfxz1UvWuMQ7EbvqwQ== =fAQW -----END PGP SIGNATURE----- --gJgclWXWHobp9Jm8TzWLGKLDSFUeHbtMu--

1 0

(ITS#9087) Minor patch for OpenLDAP use of windres when cross-compiling
by zagarc＠oclc.org 20 Sep '19

20 Sep '19

Full_Name: Chris Zagar Version: 2.4.48 OS: URL: ftp://ftp.openldap.org/incoming/chris-zagar-190920.patch Submission from: (NULL) (68.98.212.84) The file libraries/liblutil/Makefile.in is hard-coded to call windres when compiling the resources for slapdmsg.rc. This causes a problem when building using a cross-compiler. Please consider using $(RC) instead to allow the complete name for windres to be specified. Thanks!

1 0

(ITS#9086) Add debug logging for GnuTLS configuration errors
by ryan＠openldap.org 20 Sep '19

20 Sep '19

Full_Name: Ryan Tandy Version: master OS: Debian 10/buster URL: https://github.com/openldap/openldap/compare/master...rtandy:gnutls-logging… Submission from: (NULL) (70.66.128.207) Submitted by: ryan This patch adds debug logging for common GnuTLS configuration errors. It should help with troubleshooting issues such as slapd failing to start due to incorrect paths or permissions, or client TLS connections failing if the CA file can't be read. Posting here for review. I'll upload this to Debian as well and try to get some feedback from users. -- The attached patch file is derived from OpenLDAP Software. All of the modifications to OpenLDAP Software represented in the following patch(es) were developed by Ryan Tandy <ryan(a)nardis.ca>. I have not assigned rights and/or interest in this work to any party. I, Ryan Tandy, hereby place the following modifications to OpenLDAP Software (and only these modifications) into the public domain. Hence, these modifications may be freely used and/or redistributed for any purpose with or without attribution and/or other notice.

1 0

(ITS#9085) Test suite is failing
by kloczko.tomasz＠gmail.com 19 Sep '19

19 Sep '19

Full_Name: Tomasz Kloczko Version: 2.4.48 OS: Linux URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (92.25.145.71) + /usr/bin/make -O -j48 V=1 VERBOSE=1 check -j1 cd tests; /usr/bin/make test make[1]: Entering directory '/home/tkloczko/rpmbuild/BUILD/openldap-2.4.48/tests' make[2]: Entering directory '/home/tkloczko/rpmbuild/BUILD/openldap-2.4.48/tests' Initiating LDAP tests for BDB... Running ./scripts/all for bdb... >>>>> Executing all LDAP tests for bdb >>>>> Starting test000-rootdse for bdb... running defines.sh Starting slapd on TCP/IP port 9011... Using ldapsearch to retrieve the root DSE... Using ldapsearch to retrieve the cn=Subschema... Using ldapsearch to retrieve the cn=Monitor... dn: objectClass: top objectClass: OpenLDAProotDSE structuralObjectClass: OpenLDAProotDSE configContext: cn=config namingContexts: o=OpenLDAP Project,l=Internet monitorContext: cn=Monitor supportedControl: 2.16.840.1.113730.3.4.18 supportedControl: 2.16.840.1.113730.3.4.2 supportedControl: 1.3.6.1.4.1.4203.1.10.1 supportedControl: 1.3.6.1.1.22 supportedControl: 1.2.840.113556.1.4.319 supportedControl: 1.2.826.0.1.3344810.2.3 supportedControl: 1.3.6.1.1.13.2 supportedControl: 1.3.6.1.1.13.1 supportedControl: 1.3.6.1.1.12 supportedExtension: 1.3.6.1.4.1.4203.1.11.1 supportedExtension: 1.3.6.1.4.1.4203.1.11.3 supportedExtension: 1.3.6.1.1.8 supportedFeatures: 1.3.6.1.1.14 supportedFeatures: 1.3.6.1.4.1.4203.1.5.1 supportedFeatures: 1.3.6.1.4.1.4203.1.5.2 supportedFeatures: 1.3.6.1.4.1.4203.1.5.3 supportedFeatures: 1.3.6.1.4.1.4203.1.5.4 supportedFeatures: 1.3.6.1.4.1.4203.1.5.5 supportedLDAPVersion: 3 supportedSASLMechanisms: GSS-SPNEGO supportedSASLMechanisms: GSSAPI vendorName: The OpenLDAP Project <http://www.openldap.org/> entryDN: subschemaSubentry: cn=Subschema dn: cn=Subschema objectClass: top objectClass: subentry objectClass: subschema objectClass: extensibleObject cn: Subschema dn: cn=Monitor objectClass: monitorServer cn: Monitor description: This subtree contains monitoring/managing objects. description: This object contains information about this server. description: Most of the information is held in operational attributes, which must be explicitly requested. monitoredInfo: OpenLDAP: slapd 2.4.48 (Sep 20 2019 05:49:42) ./scripts/test000-rootdse: line 68: 1709599 Hangup $SLAPD -f $CONF1 -h $URI1 -d $LVL $TIMING > $LOG1 2>&1 >>>>> Test succeeded >>>>> test000-rootdse completed OK for bdb. >>>>> Starting test001-slapadd for bdb... running defines.sh Running slapadd to build slapd database... Starting slapd on TCP/IP port 9011... Using ldapsearch to retrieve all the entries... Waiting 5 seconds for slapd to start... Waiting 5 seconds for slapd to start... Waiting 5 seconds for slapd to start... Waiting 5 seconds for slapd to start... Waiting 5 seconds for slapd to start... Waiting 5 seconds for slapd to start... ldapsearch failed (32)! ./scripts/test001-slapadd: line 55: kill: (1709740) - No such process >>>>> test001-slapadd failed for bdb (exit 32) make[2]: *** [Makefile:294: bdb-yes] Error 32 make[2]: Leaving directory '/home/tkloczko/rpmbuild/BUILD/openldap-2.4.48/tests' make[1]: *** [Makefile:279: test] Error 2 make[1]: Leaving directory '/home/tkloczko/rpmbuild/BUILD/openldap-2.4.48/tests' make: *** [Makefile:291: test] Error 2 error: Bad exit status from /var/tmp/rpm-tmp.aVXfhD (%check)

1 0

Re: (ITS#9017) Improving performance of commit sync in Windows
by kriszyp＠gmail.com 18 Sep '19

18 Sep '19

--000000000000d8a73d0592d86418 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Checking on this again, is this still a possibility for merging into LMDB? This fix is still working great (improved performance) on our systems. Thanks, Kris On Mon, Jun 17, 2019 at 1:04 PM Kris Zyp <kriszyp(a)gmail.com> wrote: > Is this still being considered/reviewed? Let me know if there are any > other changes you would like me to make. This patch has continued to yiel= d > significant and reliable performance improvements for us, and seems like = it > would be nice for this to be available for other Windows users. > > On Fri, May 3, 2019 at 3:52 PM Kris Zyp <kriszyp(a)gmail.com> wrote: > >> For the sake of putting this in the email thread (other code discussion >> in GitHub), here is the latest squashed commit of the proposed patch (wi= th >> the on-demand, retained overlapped array to reduce re-malloc and opening >> event handles): >> https://github.com/kriszyp/node-lmdb/commit/726a9156662c703bf3d453aab75e= e222072b990f >> >> >> >> Thanks, >> Kris >> >> >> >> *From: *Kris Zyp <kriszyp(a)gmail.com> >> *Sent: *April 30, 2019 12:43 PM >> *To: *Howard Chu <hyc(a)symas.com>; openldap-its(a)OpenLDAP.org >> *Subject: *RE: (ITS#9017) Improving performance of commit sync in Window= s >> >> >> >> > What is the point of using writemap mode if you still need to use >> WriteFile >> >> > on every individual page? >> >> >> >> As I understood from the documentation, and have observed, using writema= p >> mode is faster (and uses less temporary memory) because it doesn=E2=80= =99t require >> mallocs to allocate pages (docs: =E2=80=9CThis is faster and uses fewer = mallocs=E2=80=9D). >> To be clear though, LMDB is so incredibly fast and efficient, that in >> sync-mode, it takes enormous transactions before the time spent allocati= ng >> and creating the dirty pages with the updated b-tree is anywhere even >> remotely close to the time it takes to wait for disk flushing, even with= an >> SSD. But the more pertinent question is efficiency, and measuring CPU >> cycles rather than time spent (efficiency is more important than just ti= me >> spent). When I ran my tests this morning of 100 (sync) transactions with >> 100 puts per transaction, times varied quite a bit, but it seemed like >> running with writemap enabled typically averages about 500ms of CPU and >> with writemap disabled it typically averages around 600ms. Not a huge >> difference, but still definitely worthwhile, I think. >> >> >> >> Caveat emptor: Measuring LMDB performance with sync interactions on >> Windows is one of the most frustratingly erratic things to measure. It i= s >> sunny outside right now, times could be different when it starts raining >> later, but, this is what I saw this morning... >> >> >> >> > What is the performance difference between your patch using writemap, >> and just >> >> > not using writemap in the first place? >> >> >> >> Running 1000 sync transactions on 3GB db with a single put per >> transaction, without writemap map, without the patch took about 60 secon= ds. >> And it took about 1 second with the patch with writemap mode enabled! >> (there is no significant difference in sync times with writemap enabled = or >> disabled with the patch.) So the difference was huge in my test. And not >> only that, without the patch, the CPU usage was actually _*higher*_ >> during that 60 seconds (close to 100% of a core) than during the executi= on >> with the patch for one second (close to 50%). Anyway, there are certain= ly >> tests I have run where the differences are not as large (doing small >> commits on large dbs accentuates the differences), but the patch always >> seems to win. It could also be that my particular configuration causes >> bigger differences (on an SSD drive, and maybe a more fragmented file?). >> >> >> >> Anyway, I added error handling for the malloc, and fixed/changed the >> other things you suggested. Be happy to make any other changes you want. >> The updated patch is here: >> >> >> https://github.com/kriszyp/node-lmdb/commit/25366dea9453749cf6637f43ec17= b9b62094acde >> >> >> >> > OVERLAPPED* ov =3D malloc((pagecount - keep) * sizeof(OVERLAPPED)); >> >> > Probably this ought to just be pre-allocated based on the maximum >> number of dirty pages a txn allows. >> >> >> >> I wasn=E2=80=99t sure I understood this comment. Are you suggesting we m= alloc(MDB_IDL_UM_MAX >> * sizeof(OVERLAPPED)) for each environment, and retain it for the life o= f >> the environment? I think that is 4MB, if my math is right, which seems l= ike >> a lot of memory to keep allocated (we usually have a lot of open >> environments). If the goal is to reduce the number of mallocs, how about= we >> retain the OVERLAPPED array, and only free and re-malloc if the previous >> allocation wasn=E2=80=99t large enough? Then there isn=E2=80=99t unneces= sary allocation, >> and we only malloc when there is a bigger transaction than any previous.= I >> put this together in a separate commit, as I wasn=E2=80=99t sure if this= what you >> wanted (can squash if you prefer): >> https://github.com/kriszyp/node-lmdb/commit/2fe68fb5269c843e2e789746a17a= 4b2adefaac40 >> >> >> >> Thank you for the review! >> >> >> >> Thanks, >> Kris >> >> >> >> *From: *Howard Chu <hyc(a)symas.com> >> *Sent: *April 30, 2019 7:12 AM >> *To: *kriszyp(a)gmail.com; openldap-its(a)OpenLDAP.org >> *Subject: *Re: (ITS#9017) Improving performance of commit sync in Window= s >> >> >> >> kriszyp(a)gmail.com wrote: >> >> > Full_Name: Kristopher William Zyp >> >> > Version: LMDB 0.9.23 >> >> > OS: Windows >> >> > URL: >> https://github.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab= 9332b7fc4ce9 >> >> > Submission from: (NULL) (71.199.6.148) >> >> > >> >> > >> >> > We have seen very poor performance on the sync of commits on large >> databases in >> >> > Windows. On databases with 2GB of data, in writemap mode, the sync of >> even small >> >> > commits is consistently well over 100ms (without writemap it is faster= , >> but >> >> > still slow). It is expected that a sync should take some time while >> waiting for >> >> > disk confirmation of the writes, but more concerning is that these syn= c >> >> > operations (in writemap mode) are instead dominated by nearly 100% >> system CPU >> >> > utilization, so operations that requires sub-millisecond b-tree update >> >> > operations are then dominated by very large amounts of system CPU >> cycles during >> >> > the sync phase. >> >> > >> >> > I think that the fundamental problem is that FlushViewOfFile seems to >> be an O(n) >> >> > operation where n is the size of the file (or map). I presume that >> Windows is >> >> > scanning the entire map/file for dirty pages to flush, I'm guessing >> because it >> >> > doesn't have an internal index of all the dirty pages for every >> file/map-view in >> >> > the OS disk cache. Therefore, the turns into an extremely expensive, >> CPU-bound >> >> > operation to find the dirty pages for (large file) and initiate their >> writes, >> >> > which, of course, is contrary to the whole goal of a scalable database >> system. >> >> > And FlushFileBuffers is also relatively slow as well. We have attempte= d >> to batch >> >> > as many operations into single transaction as possible, but this is >> still a very >> >> > large overhead. >> >> > >> >> > The Windows docs for FlushFileBuffers itself warns about the >> inefficiencies of >> >> > this function ( >> https://docs.microsoft.com/en-us/windows/desktop/api/fileapi/nf-fileapi-= flushfilebuffers >> ). >> >> > Which also points to the solution: it is much faster to write out the >> dirty >> >> > pages with WriteFile through a sync file handle >> (FILE_FLAG_WRITE_THROUGH). >> >> > >> >> > The associated patch >> >> > ( >> https://github.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab= 9332b7fc4ce9 >> ) >> >> > is my attempt at implementing this solution, for Windows. Fortunately, >> with the >> >> > design of LMDB, this is relatively straightforward. LMDB already >> supports >> >> > writing out dirty pages with WriteFile calls. I added a write-through >> handle for >> >> > sending these writes directly to disk. I then made that file-handle >> >> > overlapped/asynchronously, so all the writes for a commit could be >> started in >> >> > overlap mode, and (at least theoretically) transfer in parallel to the >> drive and >> >> > then used GetOverlappedResult to wait for the completion. So basically >> >> > mdb_page_flush becomes the sync. I extended the writing of dirty pages >> through >> >> > WriteFile to writemap mode as well (for writing meta too), so that >> WriteFile >> >> > with write-through can be used to flush the data without ever needing >> to call >> >> > FlushViewOfFile or FlushFileBuffers. I also implemented support for >> write >> >> > gathering in writemap mode where contiguous file positions infers >> contiguous >> >> > memory (by tracking the starting position with wdp and writing >> contiguous pages >> >> > in single operations). Sorting of the dirty list is maintained even in >> writemap >> >> > mode for this purpose. >> >> >> >> What is the point of using writemap mode if you still need to use >> WriteFile >> >> on every individual page? >> >> >> >> > The performance benefits of this patch, in my testing, are >> considerable. Writing >> >> > out/syncing transactions is typically over 5x faster in writemap mode, >> and 2x >> >> > faster in standard mode. And perhaps more importantly (especially in >> environment >> >> > with many threads/processes), the efficiency benefits are even larger, >> >> > particularly in writemap mode, where there can be a 50-100x reduction >> in the >> >> > system CPU usage by using this patch. This brings windows performance >> with >> >> > sync'ed transactions in LMDB back into the range of "lightning" >> performance :). >> >> >> >> What is the performance difference between your patch using writemap, an= d >> just >> >> not using writemap in the first place? >> >> >> >> -- >> >> -- Howard Chu >> >> CTO, Symas Corp. http://www.symas.com >> >> Director, Highland Sun http://highlandsun.com/hyc/ >> >> Chief Architect, OpenLDAP http://www.openldap.org/project/ >> >> >> >> >> > --000000000000d8a73d0592d86418 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr">Checking on this again, is this still a possibility for me= rging into LMDB? This fix is still working great (improved performance) on = our systems.<div>Thanks,</div><div>Kris</div></div> <div class=3D"gmail_= quote"><div dir=3D"ltr" class=3D"gmail_attr">On Mon, Jun 17, 2019 at 1:04 P= M Kris Zyp <<a href=3D"mailto:kriszyp@gmail.com">kriszyp(a)gmail.com</a>&g= t; wrote: </div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0p= x 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div d= ir=3D"ltr">Is this still being considered/reviewed? Let me know if there ar= e any other changes you would like me to make. This patch has continued to = yield significant and reliable performance improvements for us, and seems l= ike it would be nice for this to be available for other Windows users.</div= > <div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Fr= i, May 3, 2019 at 3:52 PM Kris Zyp <<a href=3D"mailto:kriszyp@gmail.com"= target=3D"_blank">kriszyp(a)gmail.com</a>> wrote: </div><blockquote cl= ass=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid= rgb(204,204,204);padding-left:1ex"><div lang=3D"EN-CA"><div class=3D"gmail= -m_-3761698270430747347gmail-m_5242437559147988140WordSection1">For the sake of putting this in the email thread (other code di= scussion in GitHub), here is the latest squashed commit of the proposed pat= ch (with the on-demand, retained overlapped array to reduce re-malloc and o= pening event handles): <a href=3D"https://github.com/kriszyp/node-lmdb/comm= it/726a9156662c703bf3d453aab75ee222072b990f" target=3D"_blank">https://gith= ub.com/kriszyp/node-lmdb/commit/726a9156662c703bf3d453aab75ee222072b990f</a= >=C2=A0= Thanks, Kris=C2=A0<div styl= e=3D"border-right:none;border-bottom:none;border-left:none;border-top:1pt s= olid rgb(225,225,225);padding:3pt 0cm 0cm">From: <a href=3D"mailto:kriszyp@gmail.com" = target=3D"_blank">Kris Zyp</a> Sent: April 30, 2019 12:43 PM <= b>To: <a href=3D"mailto:hyc@symas.com" target=3D"_blank">Howard Chu</a>= ; <a href=3D"mailto:openldap-its@OpenLDAP.org" target=3D"_blank">openldap-i= ts(a)OpenLDAP.org</a> Subject: RE: (ITS#9017) Improving performance= of commit sync in Windows</div>=C2=A0= > What is the point of using writemap mod= e if you still need to use WriteFile> on every individual page?<u= >=C2=A0As I understood from the docum= entation, and have observed, using writemap mode is faster (and uses less t= emporary memory) because it doesn=E2=80=99t require mallocs to allocate pag= es (docs: =E2=80=9CThis is faster and uses fewer mallocs=E2=80=9D). To be c= lear though, LMDB is so incredibly fast and efficient, that in sync-mode, i= t takes enormous transactions before the time spent allocating and creating= the dirty pages with the updated b-tree is anywhere even remotely close to= the time it takes to wait for disk flushing, even with an SSD. But the mor= e pertinent question is efficiency, and measuring CPU cycles rather than ti= me spent (efficiency is more important than just time spent). When I ran my= tests this morning of 100 (sync) transactions with 100 puts per transactio= n, times varied quite a bit, but it seemed like running with writemap enabl= ed typically averages about 500ms of CPU and with writemap disabled it typi= cally averages around 600ms. Not a huge difference, but still definitely wo= rthwhile, I think.=C2=A0= Caveat emptor: Measuring LMDB performance wi= th sync interactions on Windows is one of the most frustratingly erratic th= ings to measure. It is sunny outside right now, times could be different wh= en it starts raining later, but, this is what I saw this morning...<= u>=C2=A0> What is the performance difference between your patch using write= map, and just> not using writem= ap in the first place?=C2= =A0Running 1000 sync transactions on 3GB = db with a single put per transaction, without writemap map, without the pat= ch took about 60 seconds. And it took about 1 second with the patch with wr= itemap mode enabled! (there is no significant difference in sync times with= writemap enabled or disabled with the patch.) So the difference was huge i= n my test. And not only that, without the patch, the CPU usage was actually= _higher_ during that 60 seconds (close to 100% of a core) than duri= ng the execution with the patch for one second (close to 50%).=C2=A0 Anyway= , there are certainly tests I have run where the differences are not as lar= ge (doing small commits on large dbs accentuates the differences), but the = patch always seems to win. It could also be that my particular configuratio= n causes bigger differences (on an SSD drive, and maybe a more fragmented f= ile?).=C2=A0Anyway, I added error handling for the malloc, and fixed= /changed the other things you suggested. Be happy to make any other changes= you want. The updated patch is here:<a href=3D"https://github.com/kriszyp/node-lmdb/commit/25366dea9453749cf= 6637f43ec17b9b62094acde" target=3D"_blank">https://github.com/kriszyp/node-= lmdb/commit/25366dea9453749cf6637f43ec17b9b62094acde</a><= p class=3D"MsoNormal">=C2=A0><s= pan class=3D"gmail-m_-3761698270430747347gmail-m_5242437559147988140blob-co= de-inner"> OVERLAPPED* ov =3D malloc<span= style=3D"font-size:9pt;font-family:Consolas;color:rgb(36,41,46)">((pagecou= nt - keep) * sizeof(OVERLAPPED));<= /span>> Probably this o= ught to just be pre-allocated based on the maximum number of dirty pages a = txn allows.=C2=A0</span= >I wasn= =E2=80=99t sure I understood this comment. Are you suggesting we mal= loc(MDB_IDL_UM_MAX * sizeof(OVERLAPPED)) for each environment, and retain i= t for the life of the environment? I think that is 4MB, if my math is right= , which seems like a lot of memory to keep allocated (we usually have a lot= of open environments). If the goal is to reduce the number of mallocs, how= about we retain the OVERLAPPED array, and only free and re-malloc if the p= revious allocation wasn=E2=80=99t large enough? Then there isn=E2=80=99t un= necessary allocation, and we only malloc when there is a bigger transaction= than any previous. I put this together in a separate commit, as I wasn=E2= =80=99t sure if this what you wanted (can squash if you prefer): <a href=3D= "https://github.com/kriszyp/node-lmdb/commit/2fe68fb5269c843e2e789746a17a4b= 2adefaac40" target=3D"_blank">https://github.com/kriszyp/node-lmdb/commit/2= fe68fb5269c843e2e789746a17a4b2adefaac40</a>=C2=A0Thank you for the r= eview! =C2=A0Thanks,<b= r>Kris=C2=A0<div= style=3D"border-right:none;border-bottom:none;border-left:none;border-top:= 1pt solid rgb(225,225,225);padding:3pt 0cm 0cm">F= rom: <a href=3D"mailto:hyc@symas.com" target=3D"_blank">Howard Chu</a><= br>Sent: April 30, 2019 7:12 AM To: <a href=3D"mailto:kris= zyp(a)gmail.com" target=3D"_blank">kriszyp(a)gmail.com</a>; <a href=3D"mailto:o= penldap-its(a)OpenLDAP.org" target=3D"_blank">openldap-its(a)OpenLDAP.org</a><b= r>Subject: Re: (ITS#9017) Improving performance of commit sync in Wi= ndows</div>=C2=A0</p= ><a href=3D"mailto:kriszyp@gmail.com" target=3D"_bla= nk">kriszyp(a)gmail.com</a> wrote:&g= t; Full_Name: Kristopher William Zyp> Version: LMDB 0.9.23> OS= : Windows> URL: <a href=3D"http= s://github.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab9332b7f= c4ce9" target=3D"_blank">https://github.com/kriszyp/node-lmdb/commit/7ff525= ae57684a163d32af74a0ab9332b7fc4ce9</a>> Submission from: (NULL) (71.199.6.148)> > <u= >> We have seen very poor performance on = the sync of commits on large databases in> Windows. On databases with 2GB of data, in writemap mode, the s= ync of even small> commits is c= onsistently well over 100ms (without writemap it is faster, but</= u>> still slow). It is expected that a sync s= hould take some time while waiting for> disk confirmation of the writes, but more concerning is that these= sync> operations (in writemap = mode) are instead dominated by nearly 100% system CPU> utilization, so operations that requires sub-millis= econd b-tree update> operations= are then dominated by very large amounts of system CPU cycles during</u= >> the sync phase.<p= class=3D"MsoNormal">> > I t= hink that the fundamental problem is that FlushViewOfFile seems to be an O(= n)> operation where n is the si= ze of the file (or map). I presume that Windows is> scanning the entire map/file for dirty pages to flush,= I'm guessing because it> d= oesn't have an internal index of all the dirty pages for every file/map= -view in> the OS disk cache. Th= erefore, the turns into an extremely expensive, CPU-bound= > operation to find the dirty pages for (large fi= le) and initiate their writes,>= which, of course, is contrary to the whole goal of a scalable database sys= tem.> And FlushFileBuffers is a= lso relatively slow as well. We have attempted to batch<p= class=3D"MsoNormal">> as many operations into single transaction as pos= sible, but this is still a very&gt= ; large overhead.> </= u>> The Windows docs for FlushFileBuffers its= elf warns about the inefficiencies of> this function (<a href=3D"https://docs.microsoft.com/en-us/windows/= desktop/api/fileapi/nf-fileapi-flushfilebuffers" target=3D"_blank">https://= docs.microsoft.com/en-us/windows/desktop/api/fileapi/nf-fileapi-flushfilebu= ffers</a>).> Which also points = to the solution: it is much faster to write out the dirty= > pages with WriteFile through a sync file handle= (FILE_FLAG_WRITE_THROUGH).> <u= >> The associated patch<= /u>> (<a href=3D"https://github.com/kriszyp/n= ode-lmdb/commit/7ff525ae57684a163d32af74a0ab9332b7fc4ce9" target=3D"_blank"= >https://github.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab93= 32b7fc4ce9</a>)> is my attempt = at implementing this solution, for Windows. Fortunately, with the= > design of LMDB, this is relatively stra= ightforward. LMDB already supports= > writing out dirty pages with WriteFile calls. I added a write-through = handle for> sending these write= s directly to disk. I then made that file-handle> overlapped/asynchronously, so all the writes for a comm= it could be started in> overlap= mode, and (at least theoretically) transfer in parallel to the drive and<u= >> then used GetOverlappedResult t= o wait for the completion. So basically> mdb_page_flush becomes the sync. I extended the writing of dirty = pages through> WriteFile to wri= temap mode as well (for writing meta too), so that WriteFile<= /p>> with write-through can be used to flush the = data without ever needing to call&= gt; FlushViewOfFile or FlushFileBuffers. I also implemented support for wri= te> gathering in writemap mode = where contiguous file positions infers contiguous> memory (by tracking the starting position with wdp and = writing contiguous pages> in si= ngle operations). Sorting of the dirty list is maintained even in writemap<= u>> mode for this purpose.<= u>=C2=A0What is the point of using writemap mode if you still need to use Writ= eFileon every individual page?<= /u>=C2=A0> The performance benefits of this patch, in my testing, are co= nsiderable. Writing> out/syncin= g transactions is typically over 5x faster in writemap mode, and 2x<= u>> faster in standard mode. And perhaps = more importantly (especially in environment> with many threads/processes), the efficiency benefits are eve= n larger,> particularly in writ= emap mode, where there can be a 50-100x reduction in the<= p class=3D"MsoNormal">> system CPU usage by using this patch. This bring= s windows performance with> syn= c'ed transactions in LMDB back into the range of "lightning" = performance :).=C2=A0</u= >What is the performance difference between your= patch using writemap, and justnot= using writemap in the first place?=C2=A0-- =C2=A0=C2=A0-- Howard Chu=C2=A0 CTO, Symas Corp.=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 <a href=3D"http://www.symas.com" target=3D"_blank">http:= //www.symas.com</a>=C2=A0 Director= , Highland Sun=C2=A0=C2=A0=C2=A0=C2=A0 <a href=3D"http://highlandsun.com/hy= c/" target=3D"_blank">http://highlandsun.com/hyc/</a>=C2=A0 Chief Architect, OpenLDAP=C2=A0 <a href=3D"http:/= /www.openldap.org/project/" target=3D"_blank">http://www.openldap.org/proje= ct/</a>=C2=A0=C2=A0</div></div></blockquote></div> </blockquote></div> --000000000000d8a73d0592d86418--

1 0

(ITS#9084) Build issue when slaps is not included
by aixtools＠felt.demon.nl 18 Sep '19

18 Sep '19

Full_Name: Michael Felt Version: 2.4.48 OS: AIX URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (109.104.126.147) Build of 2.4.46 proceeds with no issue. When using the same command to build 2.4.48 the build stops with an undefined variable error: root@x066:[/data/prj/aixtools/src]grep LDAP_OPT_X_TLS *46/* *48/* openldap-2.4.46/CHANGES: Added libldap LDAP_OPT_X_TLS_PACKAGE (ITS#6969) openldap-2.4.48/CHANGES: Added libldap LDAP_OPT_X_TLS_PACKAGE (ITS#6969) root@x066:[/data/prj/aixtools/src]find *46 *48 | xargs grep LDAP_OPT_X_TLS_ECNAME openldap-2.4.48/include/ldap.h:#define LDAP_OPT_X_TLS_ECNAME 0x6012 openldap-2.4.48/libraries/libldap/tls2.c: case LDAP_OPT_X_TLS_ECNAME: openldap-2.4.48/libraries/libldap/tls2.c: case LDAP_OPT_X_TLS_ECNAME: openldap-2.4.48/servers/slapd/bconfig.c: case CFG_TLS_ECNAME: flag = LDAP_OPT_X_TLS_ECNAME; break; >From above, it appears that LDAP_OPT_X_TLS_ECNAME should be defined, but I question that as I get the following error when compiling... Entering subdirectory libldap make[2]: Entering directory '/data/prj/aixtools/static/openldap-2.4.48/libraries/libldap' /bin/sh ../../libtool --mode=compile xlc_r -I/opt/include -O2 -qmaxmem=-1 -I../../include -I../../../../src/openldap-2.4.48/include -I/opt/include -DLDAP_LIBRARY -c ../../../../src/openldap-2.4.48/libraries/libldap/tls2.c xlc_r -I/opt/include -O2 -qmaxmem=-1 -I../../include -I../../../../src/openldap-2.4.48/include -I/opt/include -DLDAP_LIBRARY -c ../../../../src/openldap-2.4.48/libraries/libldap/tls2.c -o tls2.o "../../../../src/openldap-2.4.48/libraries/libldap/tls2.c", line 658.14: 1506-045 (S) Undeclared identifier LDAP_OPT_X_TLS_ECNAME. "../../../../src/openldap-2.4.48/libraries/libldap/tls2.c", line 658.14: 1506-051 (S) Case expression must be a valid integral constant. "../../../../src/openldap-2.4.48/libraries/libldap/tls2.c", line 781.14: 1506-045 (S) Undeclared identifier LDAP_OPT_X_TLS_ECNAME. "../../../../src/openldap-2.4.48/libraries/libldap/tls2.c", line 781.14: 1506-051 (S) Case expression must be a valid integral constant. make[2]: *** [Makefile:413: tls2.lo] Error 1 make[2]: Leaving directory '/data/prj/aixtools/static/openldap-2.4.48/libraries/libldap' make[1]: *** [Makefile:297: all-common] Error 1 make[1]: Leaving directory '/data/prj/aixtools/static/openldap-2.4.48/libraries' make: *** [Makefile:313: all-common] Error 1 Researching further... Michael

1 0

(ITS#9083) Broken link in the list of commercial support partners
by marketing＠daasi.de 18 Sep '19

18 Sep '19

Full_Name: Romy Hoffart Version: - OS: - URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (37.24.13.226) Hey everyone, I am with the marketing department of DAASI International, one of the commercial support partners for OpenLDAP listed on your website. Unfortunately I noticed that the link to our website doesn't work anymore. Could you replace the link here (https://www.openldap.org/support/) with a new link to our website (https://daasi.de/en/)? I was in touch with Quanah from Symas before, she recommended me to submit this issue here. Thank you so much for your time & help, Romy

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

openldap-bugs