--000000000000db50ce058b89ac2f Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
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 yield 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@gmail.com wrote:
For the sake of putting this in the email thread (other code discussion i=
n
GitHub), here is the latest squashed commit of the proposed patch (with t=
he
on-demand, retained overlapped array to reduce re-malloc and opening even=
t
handles): https://github.com/kriszyp/node-lmdb/commit/726a9156662c703bf3d453aab75ee=
222072b990f
Thanks, Kris
*From: *Kris Zyp kriszyp@gmail.com *Sent: *April 30, 2019 12:43 PM *To: *Howard Chu hyc@symas.com; openldap-its@OpenLDAP.org *Subject: *RE: (ITS#9017) Improving performance of commit sync in Windows
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 writemap mode is faster (and uses less temporary memory) because it doesn=E2=80=99=
t require
mallocs to allocate pages (docs: =E2=80=9CThis is faster and uses fewer m=
allocs=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 allocatin=
g
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 tim=
e
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 is 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 second=
s.
And it took about 1 second with the patch with writemap mode enabled! (there is no significant difference in sync times with writemap enabled o=
r
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 executio=
n
with the patch for one second (close to 50%). Anyway, there are certainl=
y
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 othe=
r
things you suggested. Be happy to make any other changes you want. The updated patch is here:
https://github.com/kriszyp/node-lmdb/commit/25366dea9453749cf6637f43ec17b=
9b62094acde
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 ma=
lloc(MDB_IDL_UM_MAX
- sizeof(OVERLAPPED)) for each environment, and retain it for the life of
the environment? I think that is 4MB, if my math is right, which seems li=
ke
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 unnecess=
ary 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/2fe68fb5269c843e2e789746a17a4=
b2adefaac40
Thank you for the review!
Thanks, Kris
*From: *Howard Chu hyc@symas.com *Sent: *April 30, 2019 7:12 AM *To: *kriszyp@gmail.com; openldap-its@OpenLDAP.org *Subject: *Re: (ITS#9017) Improving performance of commit sync in Windows
kriszyp@gmail.com wrote:
Full_Name: Kristopher William Zyp
Version: LMDB 0.9.23
OS: Windows
URL:
https://github.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab9=
332b7fc4ce9
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 sync
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 cycle=
s
during
the sync phase.
I think that the fundamental problem is that FlushViewOfFile seems to b=
e
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 attempted
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-f=
lushfilebuffers
).
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/7ff525ae57684a163d32af74a0ab9=
332b7fc4ce9
)
is my attempt at implementing this solution, for Windows. Fortunately,
with the
design of LMDB, this is relatively straightforward. LMDB already suppor=
ts
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 t=
o
call
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 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 WriteFi=
le
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 i=
n
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, and 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/
--000000000000db50ce058b89ac2f Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
<div dir=3D"ltr">Is this still being considered/reviewed? Let me know if th= ere are any other changes you would like me to make. This patch has continu= ed to yield significant and reliable performance improvements for us, and s= eems like it would be nice for this to be available for other Windows users= .</div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr"=
On Fri, May 3, 2019 at 3:52 PM Kris Zyp <<a href=3D"mailto:kriszyp@gmai=
l.com">kriszyp@gmail.com</a>> wrote:<br></div><blockquote class=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_524243755= 9147988140WordSection1"><p class=3D"MsoNormal">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 (with the on-demand, retained overlap= ped array to reduce re-malloc and opening event handles): <a href=3D"https:= //github.com/kriszyp/node-lmdb/commit/726a9156662c703bf3d453aab75ee222072b9= 90f" target=3D"_blank">https://github.com/kriszyp/node-lmdb/commit/726a9156= 662c703bf3d453aab75ee222072b990f</a></p><p class=3D"MsoNormal"><u></u>=C2= =A0<u></u></p><p class=3D"MsoNormal">Thanks,<br>Kris</p><p class=3D"MsoNorm= al"><u></u>=C2=A0<u></u></p><div style=3D"border-right:none;border-bottom:n= one;border-left:none;border-top:1pt solid rgb(225,225,225);padding:3pt 0cm = 0cm"><p class=3D"MsoNormal" style=3D"border:none;padding:0cm"><b>From: </b>= <a href=3D"mailto:kriszyp@gmail.com" target=3D"_blank">Kris Zyp</a><br><b>S= ent: </b>April 30, 2019 12:43 PM<br><b>To: </b><a href=3D"mailto:hyc@symas.= com" target=3D"_blank">Howard Chu</a>; <a href=3D"mailto:openldap-its@OpenL= DAP.org" target=3D"_blank">openldap-its@OpenLDAP.org</a><br><b>Subject: </b=
RE: (ITS#9017) Improving performance of commit sync in Windows</p></div><p=
class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">> Wh= at is the point of using writemap mode if you still need to use WriteFile<u=
</u><u></u></p><p class=3D"MsoNormal">> on every individual page?<u></u= <u></u></p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoN=
ormal">As I understood from the documentation, and have observed, using wri= temap 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 an= d uses fewer mallocs=E2=80=9D). To be clear though, LMDB is so incredibly f= ast and efficient, that in sync-mode, it 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 more pertinent question is efficiency, = and measuring CPU cycles rather than time spent (efficiency is more importa= nt than just time spent). When I ran my tests this morning of 100 (sync) tr= ansactions 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 h= uge difference, but still definitely worthwhile, I think.<u></u><u></u></p>= <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">Cavea= t emptor: Measuring LMDB performance with sync interactions on Windows is o= ne of the most frustratingly erratic things to measure. It is sunny outside= right now, times could be different when it starts raining later, but, thi= s is what I saw this morning...<u></u><u></u></p><p class=3D"MsoNormal"><u>= </u>=C2=A0<u></u></p><p class=3D"MsoNormal">> What is the performance di= fference between your patch using writemap, and just<u></u><u></u></p><p cl= ass=3D"MsoNormal">> not using writemap in the first place?<u></u><u></u>= </p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">R= unning 1000 sync transactions on 3GB db with a single put per transaction, = without writemap map, without the patch took about 60 seconds. And it took = about 1 second with the patch with writemap mode enabled! (there is no sign= ificant 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 _<i>higher</i>_ during that 60 second= s (close to 100% of a core) than during the execution with the patch for on= e second (close to 50%).=C2=A0 Anyway, there are certainly tests I have run= where the differences are not as large (doing small commits on large dbs a= ccentuates the differences), but the patch always seems to win. It could al= so be that my particular configuration causes bigger differences (on an SSD= drive, and maybe a more fragmented file?).<u></u><u></u></p><p class=3D"Ms= oNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">Anyway, I added err= or handling for the malloc, and fixed/changed the other things you suggeste= d. Be happy to make any other changes you want. The updated patch is here:<= u></u><u></u></p><p class=3D"MsoNormal"><a href=3D"https://github.com/krisz= yp/node-lmdb/commit/25366dea9453749cf6637f43ec17b9b62094acde" target=3D"_bl= ank">https://github.com/kriszyp/node-lmdb/commit/25366dea9453749cf6637f43ec= 17b9b62094acde</a><u></u><u></u></p><p class=3D"MsoNormal"><u></u>=C2=A0<u>= </u></p><p class=3D"MsoNormal">><span class=3D"gmail-m_52424375591479881= 40blob-code-inner"><span style=3D"font-size:9pt;font-family:Consolas;color:= rgb(36,41,46)"> OVERLAPPED* ov =3D </span></span><span class=3D"gmail-m_524= 2437559147988140pl-c1"><span style=3D"font-size:9pt;font-family:Consolas;co= lor:rgb(0,92,197)">malloc</span></span><span class=3D"gmail-m_5242437559147= 988140blob-code-inner"><span style=3D"font-size:9pt;font-family:Consolas;co= lor:rgb(36,41,46)">((pagecount - keep) * </span></span><span class=3D"gmail= -m_5242437559147988140pl-k"><span style=3D"font-size:9pt;font-family:Consol= as;color:rgb(215,58,73)">sizeof</span></span><span class=3D"gmail-m_5242437= 559147988140blob-code-inner"><span style=3D"font-size:9pt;font-family:Conso= las;color:rgb(36,41,46)">(OVERLAPPED));</span></span><span class=3D"gmail-m= _5242437559147988140blob-code-inner"><span style=3D"font-size:9pt;font-fami= ly:Consolas;color:rgb(36,41,46)"><u></u><u></u></span></span></p><p class= =3D"MsoNormal"><span class=3D"gmail-m_5242437559147988140blob-code-inner"><= span style=3D"font-size:9pt;font-family:Consolas;color:rgb(36,41,46)">> = </span></span><span style=3D"font-size:10.5pt;font-family:"Segoe UI&qu= ot;,sans-serif;color:rgb(36,41,46);background:white">Probably this ought to= just be pre-allocated based on the maximum number of dirty pages a txn all= ows.</span><span style=3D"font-size:10.5pt;font-family:"Segoe UI"= ,sans-serif;background:white"><u></u><u></u></span></p><p class=3D"MsoNorma= l"><span style=3D"font-size:10.5pt;font-family:"Segoe UI",sans-se= rif;color:rgb(36,41,46);background:white"><u></u>=C2=A0<u></u></span></p><p= class=3D"MsoNormal"><span style=3D"font-size:10.5pt;font-family:"Sego= e UI",sans-serif;color:rgb(36,41,46);background:white">I wasn=E2=80=99= t sure I understood this comment. Are you suggesting we </span>malloc(MDB_I= DL_UM_MAX * sizeof(OVERLAPPED)) for each environment, and retain it for the= life of the environment? I think that is 4MB, if my math is right, which s= eems 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 previous a= llocation wasn=E2=80=99t large enough? Then there isn=E2=80=99t unnecessary= 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 su= re if this what you wanted (can squash if you prefer): <a href=3D"https://g= ithub.com/kriszyp/node-lmdb/commit/2fe68fb5269c843e2e789746a17a4b2adefaac40= " target=3D"_blank">https://github.com/kriszyp/node-lmdb/commit/2fe68fb5269= c843e2e789746a17a4b2adefaac40</a><u></u><u></u></p><p class=3D"MsoNormal"><= u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">Thank you for the review! <sp= an style=3D"font-size:10.5pt;font-family:"Segoe UI",sans-serif;co= lor:rgb(36,41,46);background:white"><u></u><u></u></span></p><p class=3D"Ms= oNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">Thanks,<br>Kris<u><= /u><u></u></p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div style=3D"= border-right:none;border-bottom:none;border-left:none;border-top:1pt solid = rgb(225,225,225);padding:3pt 0cm 0cm"><p class=3D"MsoNormal"><b>From: </b><= a href=3D"mailto:hyc@symas.com" target=3D"_blank">Howard Chu</a><br><b>Sent= : </b>April 30, 2019 7:12 AM<br><b>To: </b><a href=3D"mailto:kriszyp@gmail.= com" target=3D"_blank">kriszyp@gmail.com</a>; <a href=3D"mailto:openldap-it= s@OpenLDAP.org" target=3D"_blank">openldap-its@OpenLDAP.org</a><br><b>Subje= ct: </b>Re: (ITS#9017) Improving performance of commit sync in Windows<u></= u><u></u></p></div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class= =3D"MsoNormal"><a href=3D"mailto:kriszyp@gmail.com" target=3D"_blank">krisz= yp@gmail.com</a> wrote:<u></u><u></u></p><p class=3D"MsoNormal">> Full_N= ame: Kristopher William Zyp<u></u><u></u></p><p class=3D"MsoNormal">> Ve= rsion: LMDB 0.9.23<u></u><u></u></p><p class=3D"MsoNormal">> OS: Windows= <u></u><u></u></p><p class=3D"MsoNormal">> URL: <a href=3D"https://githu= b.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab9332b7fc4ce9" ta= rget=3D"_blank">https://github.com/kriszyp/node-lmdb/commit/7ff525ae57684a1= 63d32af74a0ab9332b7fc4ce9</a><u></u><u></u></p><p class=3D"MsoNormal">> = Submission from: (NULL) (71.199.6.148)<u></u><u></u></p><p class=3D"MsoNorm= al">> <u></u><u></u></p><p class=3D"MsoNormal">> <u></u><u></u></p><p= class=3D"MsoNormal">> We have seen very poor performance on the sync of= commits on large databases in<u></u><u></u></p><p class=3D"MsoNormal">>= Windows. On databases with 2GB of data, in writemap mode, the sync of even= small<u></u><u></u></p><p class=3D"MsoNormal">> commits is consistently= well over 100ms (without writemap it is faster, but<u></u><u></u></p><p cl= ass=3D"MsoNormal">> still slow). It is expected that a sync should take = some time while waiting for<u></u><u></u></p><p class=3D"MsoNormal">> di= sk confirmation of the writes, but more concerning is that these sync<u></u=
<u></u></p><p class=3D"MsoNormal">> operations (in writemap mode) are i=
nstead dominated by nearly 100% system CPU<u></u><u></u></p><p class=3D"Mso= Normal">> utilization, so operations that requires sub-millisecond b-tre= e update<u></u><u></u></p><p class=3D"MsoNormal">> operations are then d= ominated by very large amounts of system CPU cycles during<u></u><u></u></p=
<p class=3D"MsoNormal">> the sync phase.<u></u><u></u></p><p class=3D"M=
soNormal">> <u></u><u></u></p><p class=3D"MsoNormal">> I think that t= he fundamental problem is that FlushViewOfFile seems to be an O(n)<u></u><u=
</u></p><p class=3D"MsoNormal">> operation where n is the size of the f=
ile (or map). I presume that Windows is<u></u><u></u></p><p class=3D"MsoNor= mal">> scanning the entire map/file for dirty pages to flush, I'm gu= essing because it<u></u><u></u></p><p class=3D"MsoNormal">> doesn't = have an internal index of all the dirty pages for every file/map-view in<u>= </u><u></u></p><p class=3D"MsoNormal">> the OS disk cache. Therefore, th= e turns into an extremely expensive, CPU-bound<u></u><u></u></p><p class=3D= "MsoNormal">> operation to find the dirty pages for (large file) and ini= tiate their writes,<u></u><u></u></p><p class=3D"MsoNormal">> which, of = course, is contrary to the whole goal of a scalable database system.<u></u>= <u></u></p><p class=3D"MsoNormal">> And FlushFileBuffers is also relativ= ely slow as well. We have attempted to batch<u></u><u></u></p><p class=3D"M= soNormal">> as many operations into single transaction as possible, but = this is still a very<u></u><u></u></p><p class=3D"MsoNormal">> large ove= rhead.<u></u><u></u></p><p class=3D"MsoNormal">> <u></u><u></u></p><p cl= ass=3D"MsoNormal">> The Windows docs for FlushFileBuffers itself warns a= bout the inefficiencies of<u></u><u></u></p><p class=3D"MsoNormal">> thi= s function (<a href=3D"https://docs.microsoft.com/en-us/windows/desktop/api= /fileapi/nf-fileapi-flushfilebuffers" target=3D"_blank">https://docs.micros= oft.com/en-us/windows/desktop/api/fileapi/nf-fileapi-flushfilebuffers</a>).= <u></u><u></u></p><p class=3D"MsoNormal">> Which also points to the solu= tion: it is much faster to write out the dirty<u></u><u></u></p><p class=3D= "MsoNormal">> pages with WriteFile through a sync file handle (FILE_FLAG= _WRITE_THROUGH).<u></u><u></u></p><p class=3D"MsoNormal">> <u></u><u></u=
</p><p class=3D"MsoNormal">> The associated patch<u></u><u></u></p><p c=
lass=3D"MsoNormal">> (<a href=3D"https://github.com/kriszyp/node-lmdb/co= mmit/7ff525ae57684a163d32af74a0ab9332b7fc4ce9" target=3D"_blank">https://gi= thub.com/kriszyp/node-lmdb/commit/7ff525ae57684a163d32af74a0ab9332b7fc4ce9<= /a>)<u></u><u></u></p><p class=3D"MsoNormal">> is my attempt at implemen= ting this solution, for Windows. Fortunately, with the<u></u><u></u></p><p = class=3D"MsoNormal">> design of LMDB, this is relatively straightforward= . LMDB already supports<u></u><u></u></p><p class=3D"MsoNormal">> writin= g out dirty pages with WriteFile calls. I added a write-through handle for<= u></u><u></u></p><p class=3D"MsoNormal">> sending these writes directly = to disk. I then made that file-handle<u></u><u></u></p><p class=3D"MsoNorma= l">> overlapped/asynchronously, so all the writes for a commit could be = started in<u></u><u></u></p><p class=3D"MsoNormal">> overlap mode, and (= at least theoretically) transfer in parallel to the drive and<u></u><u></u>= </p><p class=3D"MsoNormal">> then used GetOverlappedResult to wait for t= he completion. So basically<u></u><u></u></p><p class=3D"MsoNormal">> md= b_page_flush becomes the sync. I extended the writing of dirty pages throug= h<u></u><u></u></p><p class=3D"MsoNormal">> WriteFile to writemap mode a= s well (for writing meta too), so that WriteFile<u></u><u></u></p><p class= =3D"MsoNormal">> with write-through can be used to flush the data withou= t ever needing to call<u></u><u></u></p><p class=3D"MsoNormal">> FlushVi= ewOfFile or FlushFileBuffers. I also implemented support for write<u></u><u=
</u></p><p class=3D"MsoNormal">> gathering in writemap mode where conti=
guous file positions infers contiguous<u></u><u></u></p><p class=3D"MsoNorm= al">> memory (by tracking the starting position with wdp and writing con= tiguous pages<u></u><u></u></p><p class=3D"MsoNormal">> in single operat= ions). Sorting of the dirty list is maintained even in writemap<u></u><u></= u></p><p class=3D"MsoNormal">> mode for this purpose.<u></u><u></u></p><= p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">What i= s the point of using writemap mode if you still need to use WriteFile<u></u=
<u></u></p><p class=3D"MsoNormal">on every individual page?<u></u><u></u><=
/p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">&g= t; The performance benefits of this patch, in my testing, are considerable.= Writing<u></u><u></u></p><p class=3D"MsoNormal">> out/syncing transacti= ons is typically over 5x faster in writemap mode, and 2x<u></u><u></u></p><= p class=3D"MsoNormal">> faster in standard mode. And perhaps more import= antly (especially in environment<u></u><u></u></p><p class=3D"MsoNormal">&g= t; with many threads/processes), the efficiency benefits are even larger,<u=
</u><u></u></p><p class=3D"MsoNormal">> particularly in writemap mode, =
where there can be a 50-100x reduction in the<u></u><u></u></p><p class=3D"= MsoNormal">> system CPU usage by using this patch. This brings windows p= erformance with<u></u><u></u></p><p class=3D"MsoNormal">> sync'ed tr= ansactions in LMDB back into the range of "lightning" performance= :).<u></u><u></u></p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p cla= ss=3D"MsoNormal">What is the performance difference between your patch usin= g writemap, and just<u></u><u></u></p><p class=3D"MsoNormal">not using writ= emap in the first place?<u></u><u></u></p><p class=3D"MsoNormal"><u></u>=C2= =A0<u></u></p><p class=3D"MsoNormal">-- <u></u><u></u></p><p class=3D"MsoNo= rmal">=C2=A0=C2=A0-- Howard Chu<u></u><u></u></p><p class=3D"MsoNormal">=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><u></u><u></u></p><p class=3D"MsoNormal">=C2=A0 Director, Highland S= un=C2=A0=C2=A0=C2=A0=C2=A0 <a href=3D"http://highlandsun.com/hyc/" target= =3D"_blank">http://highlandsun.com/hyc/</a><u></u><u></u></p><p class=3D"Ms= oNormal">=C2=A0 Chief Architect, OpenLDAP=C2=A0 <a href=3D"http://www.openl= dap.org/project/" target=3D"_blank">http://www.openldap.org/project/</a><u>= </u><u></u></p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"M= soNormal"><u></u>=C2=A0<u></u></p></div></div></blockquote></div>
--000000000000db50ce058b89ac2f--