Обсуждение: Shared Buffer Size

Hi All,

I am little confused about the internal working of PostgreSQL. There is a parameter shared_buffer in postgres.conf and I am assuming that it is used for buffer management in PostgreSQL. If there is a need to bring in a new page in the buffer and size exceeds the shared_buffer limit, a victim dirty page will be written back to the disk.

However, I have read on many links that PostgreSQL depends on the OS for caching. (http://www.varlena.com/GeneralBits/Tidbits/perf.html#shbuf)

So my question is, the actual limit of the shared buffer will be defined by OS or the shared_buffer parameter in the postgres.conf to figure whether a victim dirty page needs to be selected for disk write or not?

Thanks!

Check out the "Inside the PostgreSQL Buffer Cache" link here:

http://projects.2ndquadrant.com/talks

Thanks to Greg Smith (active here).

Derrick

Hi Derrick,

Hi Preetika,

Hi Carl,

On 05/28/2011 04:42 AM, Carl von Clausewitz wrote:
> I've try to get any information about optimal memory config, and
> working, but there wasn't any "optimal memory setting calculator" on
> the internet, just some guide in the posgre documentation

There's a simple one at https://github.com/gregs1104/pgtune and the
guide at http://wiki.postgresql.org/wiki/Tuning_Your_PostgreSQL_Server
covers many of the most popular questions here too.  The right
proportion of memory setting is very dependent on workload, which makes
any sort of calculator hard to create.  pgtune takes a workload type as
an input to help with that, but the settings that come out should be
considered starting values only.  You'll need to monitoring how much
memory is actually being used by the server, as well as the output from
parameters like log_time_files, to know for sure if things are working well.

--
Greg Smith   2ndQuadrant US    greg@2ndQuadrant.com   Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support  www.2ndQuadrant.us
"PostgreSQL 9.0 High Performance": http://www.2ndQuadrant.com/books

On 05/27/2011 05:33 PM, preetika tyagi wrote:
> Hows does the shared buffer in Postgres rely on the Operating System
> cache?
> Suppose my RAM is 8 GB and shared_buffer is 24 MB in postgres. And
> there are some dirty pages in shared_buffer and I need to write a
> dirty page back to the disk to bring in a new page. What happens in
> this case? The dirty page will be written to the disk considering the
> shared_buffer size as 24 MB? or it will not be written and will stay
> in RAM which is 8 GB?
>

In this case, the dirty page will be written to the operating system,
which will initially store it in its own write cache.  Eventually the
operating system will write that page out to disk from that cache.  The
delay before that happens can easily be 30 seconds or more on operating
systems like Linux.  If PostgreSQL needs to read that data again, it may
find it still in the OS cache, in which case the read will happen very
quickly.  Eventually, the dirty data will be written to disk, and if
it's not used for long enough the OS cache memory will be re-used for
something else.

When you read a page into the database, and you don't dirty it, it might
be evicted from the database cache without a write.  If the database
needs that page again, it will ask the OS for it.  If the OS still has
it in its own read cache, it may just read it from the cache again,
without a real disk read happening.

--
Greg Smith   2ndQuadrant US    greg@2ndQuadrant.com   Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support  www.2ndQuadrant.us
"PostgreSQL 9.0 High Performance": http://www.2ndQuadrant.com/books

On 28/05/11 18:42, Carl von Clausewitz wrote:
> a few months ago, when I installed my first PostgreSQL, I have had the
> same problem. I've try to get any information about optimal memory
> config, and working, but there wasn't any "optimal memory setting
> calculator" on the internet, just some guide in the posgre documentation
> (http://www.postgresql.org/docs/9.0/interactive/kernel-resources.html#SYSVIPC).
> I got FreeBSD 8.2 AMD64, with 8 GB of memory (this server is just for
> PostgreSQL and a little PHP app with 2 user), and I have theese setting
> in postgresql.conf (which are not the default):
>
[snip]
> work_mem = 64MB# min 64kB
> maintenance_work_mem = 1024MB# min 1MB
> max_stack_depth = 64MB# min 100kB

Just a warning - but be careful about setting work_mem to high values.
The actual memory used by a query can be many times the value, depending
on the complexity of your query.

In a particular query I saw last week, we were regularly exceeding the
available memory on a server, because the query was requiring 80 times
the value of work_mem, and work_mem had been set to a high value.

Reducing work_mem back to just 4MB reduced memory usage by a couple of
gigabytes, and had almost no effect on the execution time. (Actually, it
was marginally faster - probably because more memory was left for the
operating system's cache)

Toby

Thanks Toby, I will check it, and change it. 

2011/5/30 Toby Corkindale <toby.corkindale@strategicdata.com.au>:
> On 28/05/11 18:42, Carl von Clausewitz wrote:
>>
>> a few months ago, when I installed my first PostgreSQL, I have had the
>> same problem. I've try to get any information about optimal memory
>> config, and working, but there wasn't any "optimal memory setting
>> calculator" on the internet, just some guide in the posgre documentation
>>
>> (http://www.postgresql.org/docs/9.0/interactive/kernel-resources.html#SYSVIPC).
>> I got FreeBSD 8.2 AMD64, with 8 GB of memory (this server is just for
>> PostgreSQL and a little PHP app with 2 user), and I have theese setting
>> in postgresql.conf (which are not the default):
>>
> [snip]
>>
>> work_mem = 64MB# min 64kB
>> maintenance_work_mem = 1024MB# min 1MB
>> max_stack_depth = 64MB# min 100kB
>
> Just a warning - but be careful about setting work_mem to high values.
> The actual memory used by a query can be many times the value, depending on
> the complexity of your query.
>
> In a particular query I saw last week, we were regularly exceeding the
> available memory on a server, because the query was requiring 80 times the
> value of work_mem, and work_mem had been set to a high value.
>
> Reducing work_mem back to just 4MB reduced memory usage by a couple of
> gigabytes, and had almost no effect on the execution time. (Actually, it was
> marginally faster - probably because more memory was left for the operating
> system's cache)

Maybe, you're also aware that linux may decide to swap to protect its
buffer cache (depend of the strategy it got in its configuration) and
also that you may be limited by commitable memory. On a default
install where the swap is NOT at least twice the RAM size, you're not
able to commit all RAM you have. But, it protects the buffer cache for
the not allocatable memory.

So maybe you've hitten a step where you did swap your work_mem...
anyway interesting to have a query where a large work_mem is not
better... Will it be hard to isolate the case and make it public ? In
the long term it might be a good test to add to a performance farm if
it is not based on a non-optimum linux configuration (I mean if the
issue *need* the work_mem to be reduced to be fixed).

>
> Toby
>
> --
> Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-general
>



--
Cédric Villemain               2ndQuadrant
http://2ndQuadrant.fr/     PostgreSQL : Expertise, Formation et Support

On 30/05/11 20:41, Cédric Villemain wrote:
> 2011/5/30 Toby Corkindale<toby.corkindale@strategicdata.com.au>:
>> On 28/05/11 18:42, Carl von Clausewitz wrote:
>>>
>>> a few months ago, when I installed my first PostgreSQL, I have had the
>>> same problem. I've try to get any information about optimal memory
>>> config, and working, but there wasn't any "optimal memory setting
>>> calculator" on the internet, just some guide in the posgre documentation
>>>
>>> (http://www.postgresql.org/docs/9.0/interactive/kernel-resources.html#SYSVIPC).
>>> I got FreeBSD 8.2 AMD64, with 8 GB of memory (this server is just for
>>> PostgreSQL and a little PHP app with 2 user), and I have theese setting
>>> in postgresql.conf (which are not the default):
>>>
>> [snip]
>>>
>>> work_mem = 64MB# min 64kB
>>> maintenance_work_mem = 1024MB# min 1MB
>>> max_stack_depth = 64MB# min 100kB
>>
>> Just a warning - but be careful about setting work_mem to high values.
>> The actual memory used by a query can be many times the value, depending on
>> the complexity of your query.
>>
>> In a particular query I saw last week, we were regularly exceeding the
>> available memory on a server, because the query was requiring 80 times the
>> value of work_mem, and work_mem had been set to a high value.
>>
>> Reducing work_mem back to just 4MB reduced memory usage by a couple of
>> gigabytes, and had almost no effect on the execution time. (Actually, it was
>> marginally faster - probably because more memory was left for the operating
>> system's cache)
>
> Maybe, you're also aware that linux may decide to swap to protect its
> buffer cache (depend of the strategy it got in its configuration) and
> also that you may be limited by commitable memory. On a default
> install where the swap is NOT at least twice the RAM size, you're not
> able to commit all RAM you have. But, it protects the buffer cache for
> the not allocatable memory.
>
> So maybe you've hitten a step where you did swap your work_mem...
> anyway interesting to have a query where a large work_mem is not
> better... Will it be hard to isolate the case and make it public ? In
> the long term it might be a good test to add to a performance farm if
> it is not based on a non-optimum linux configuration (I mean if the
> issue *need* the work_mem to be reduced to be fixed).


In this case, it was not just slowing down due to the amount of work_mem
allocated -- it was exceeding several gigabytes of memory usage and
crashing out. Lower values of work_mem allowed the query to succeed, but
it used almost 3G.. Even lower values of work_mem could do the query in
only a few hundred MB - and was faster.

I note that if you exceed work_mem in a query,then I guess the temp
files created are cached by the VM cache, so it's not like the
performance hit will be *too* bad?


I agree that the slowness of the 3GB version could be due to swapping or
something like that.. or just due to the VM cache being eliminated as I
suggested.

Either way - the problem was that this (machine-generated) query was
pivoting and joining many views-of-views. It's a pretty nasty query.

The key fact is that postgres (8.3) seems to allocate the full work_mem
amount every time it needs *some* work_mem - even if it could have
happily got by on just a few MB. So if your query allocates work_mem a
hundred times, it'll consume $work_mem * 100 -- or die trying.

I'm curious to know if Postgres 9.0 has improved this -- I'm going to
try re-running this query on it once I get a chance, but due to
contractual agreements this isn't quite as simple to test as you might
think.
(And running the test over a much smaller example data set might not
trigger the same query plan)
I'll get there eventually though :)

2011/5/31 Toby Corkindale <toby.corkindale@strategicdata.com.au>:
> On 30/05/11 20:41, Cédric Villemain wrote:
>>
>> 2011/5/30 Toby Corkindale<toby.corkindale@strategicdata.com.au>:
>>>
>>> On 28/05/11 18:42, Carl von Clausewitz wrote:
>>>>
>>>> a few months ago, when I installed my first PostgreSQL, I have had the
>>>> same problem. I've try to get any information about optimal memory
>>>> config, and working, but there wasn't any "optimal memory setting
>>>> calculator" on the internet, just some guide in the posgre documentation
>>>>
>>>>
>>>> (http://www.postgresql.org/docs/9.0/interactive/kernel-resources.html#SYSVIPC).
>>>> I got FreeBSD 8.2 AMD64, with 8 GB of memory (this server is just for
>>>> PostgreSQL and a little PHP app with 2 user), and I have theese setting
>>>> in postgresql.conf (which are not the default):
>>>>
>>> [snip]
>>>>
>>>> work_mem = 64MB# min 64kB
>>>> maintenance_work_mem = 1024MB# min 1MB
>>>> max_stack_depth = 64MB# min 100kB
>>>
>>> Just a warning - but be careful about setting work_mem to high values.
>>> The actual memory used by a query can be many times the value, depending
>>> on
>>> the complexity of your query.
>>>
>>> In a particular query I saw last week, we were regularly exceeding the
>>> available memory on a server, because the query was requiring 80 times
>>> the
>>> value of work_mem, and work_mem had been set to a high value.
>>>
>>> Reducing work_mem back to just 4MB reduced memory usage by a couple of
>>> gigabytes, and had almost no effect on the execution time. (Actually, it
>>> was
>>> marginally faster - probably because more memory was left for the
>>> operating
>>> system's cache)
>>
>> Maybe, you're also aware that linux may decide to swap to protect its
>> buffer cache (depend of the strategy it got in its configuration) and
>> also that you may be limited by commitable memory. On a default
>> install where the swap is NOT at least twice the RAM size, you're not
>> able to commit all RAM you have. But, it protects the buffer cache for
>> the not allocatable memory.
>>
>> So maybe you've hitten a step where you did swap your work_mem...
>> anyway interesting to have a query where a large work_mem is not
>> better... Will it be hard to isolate the case and make it public ? In
>> the long term it might be a good test to add to a performance farm if
>> it is not based on a non-optimum linux configuration (I mean if the
>> issue *need* the work_mem to be reduced to be fixed).
>
>
> In this case, it was not just slowing down due to the amount of work_mem
> allocated -- it was exceeding several gigabytes of memory usage and crashing
> out. Lower values of work_mem allowed the query to succeed, but it used
> almost 3G.. Even lower values of work_mem could do the query in only a few
> hundred MB - and was faster.
>
> I note that if you exceed work_mem in a query,then I guess the temp files
> created are cached by the VM cache, so it's not like the performance hit
> will be *too* bad?

correct.

>
>
> I agree that the slowness of the 3GB version could be due to swapping or
> something like that.. or just due to the VM cache being eliminated as I
> suggested.
>
> Either way - the problem was that this (machine-generated) query was
> pivoting and joining many views-of-views. It's a pretty nasty query.
>
> The key fact is that postgres (8.3) seems to allocate the full work_mem
> amount every time it needs *some* work_mem - even if it could have happily
> got by on just a few MB. So if your query allocates work_mem a hundred
> times, it'll consume $work_mem * 100 -- or die trying.

work_mem is not allocated fully from the beginning. It is allocated
initialy at XXKB (or MB), then doubled each time we are near the end
of the allocated memory. (I am unsure of detail like when exactly we
alloc more mem but the logic is this one)

>
> I'm curious to know if Postgres 9.0 has improved this -- I'm going to try
> re-running this query on it once I get a chance, but due to contractual
> agreements this isn't quite as simple to test as you might think.
> (And running the test over a much smaller example data set might not trigger
> the same query plan)
> I'll get there eventually though :)
>

9.0 may help you for generated queries so it is worth testing it I think.


--
Cédric Villemain               2ndQuadrant
http://2ndQuadrant.fr/     PostgreSQL : Expertise, Formation et Support

2011/5/28 Carl von Clausewitz <clausewitz45@gmail.com>:
> Hi Preetika,
> a few months ago, when I installed my first PostgreSQL, I have had the same
> problem. I've try to get any information about optimal memory config, and
> working, but there wasn't any "optimal memory setting calculator" on the
> internet, just some guide in the posgre documentation
> (http://www.postgresql.org/docs/9.0/interactive/kernel-resources.html#SYSVIPC).
> I got FreeBSD 8.2 AMD64, with 8 GB of memory (this server is just for
> PostgreSQL and a little PHP app with 2 user), and I have theese setting in
> postgresql.conf (which are not the default):
> listen_addresses = '192.168.1.1' # what IP address(es) to listen on;
> port = 5432 # (change requires restart)
> max_connections = 200 # (There are 20 user, with Microsoft Access client and
> ODBC connections... (min 6 connection / user))
> shared_buffers = 1900MB # min 128kB
> temp_buffers = 64MB # min 800kB
> work_mem = 64MB # min 64kB
> maintenance_work_mem = 1024MB # min 1MB
> max_stack_depth = 64MB # min 100kB

this max_stack_depth is unsane it should be the result of (ulimit -u)
- 1MB maximum.
Except if you have a kernel build with a stack_depth of 64MB, which
would surprised me. (common kernel have 8MB or 16MB of stack_depth)

> shared_preload_libraries = '$libdir/plpgsql.so' # (change requires restart)
> checkpoint_segments = 32 # in logfile segments, min 1, 16MB each
> checkpoint_timeout = 15min # range 30s-1h
> checkpoint_completion_target = 0.5 # checkpoint target duration, 0.0 - 1.0
> effective_cache_size = 4096MB
> track_activities = on
> track_counts = on
> #track_functions = none # none, pl, all
> #track_activity_query_size = 1024 # (change requires restart)
> update_process_title = off
> deadlock_timeout = 1s
> max_locks_per_transaction = 256 # min 10
> And the sysctl.conf from BSD, which are relevant for theese postgre
> settings:
> kern.ipc.shmall=524288
> kern.ipc.shmmax=2147483648
> kern.ipc.semmap=512
> kern.ipc.shm_use_phys=1
> And the last one is the loader.conf from BSD, which are relevant for theese
> postgre settings:
> kern.ipc.semmni=512
> kern.ipc.semmns=1024
> kern.ipc.semmnu=512
> Theese settings based on my experience, with lot of reboot and restart and
> reload config - I hope this can help you, and I accept any comment, if I
> need to set everything else :-)
> Thanks,
> Carl
> 2011/5/27 preetika tyagi <preetikatyagi@gmail.com>
>>
>> Hi Derrick,
>> Thank you for your response.
>> I saw this document and trying to understand "Interaction with the
>> Operating System Cache" which is mentioned in this document.
>> I have the following question-
>> Hows does the shared buffer in Postgres rely on the Operating System
>> cache?
>> Suppose my RAM is 8 GB and shared_buffer is 24 MB in postgres. And there
>> are some dirty pages in shared_buffer and I need to write a dirty page back
>> to the disk to bring in a new page. What happens in this case? The dirty
>> page will be written to the disk considering the shared_buffer size as 24
>> MB? or it will not be written and will stay in RAM which is 8 GB?
>> Thanks,
>> Preetika
>>
>> On Fri, May 27, 2011 at 2:11 PM, Derrick Rice <derrick.rice@gmail.com>
>> wrote:
>>>
>>> Check out the "Inside the PostgreSQL Buffer Cache" link here:
>>>
>>> http://projects.2ndquadrant.com/talks
>>>
>>> Thanks to Greg Smith (active here).
>>>
>>> Derrick
>>>
>>> On Fri, May 27, 2011 at 3:36 PM, preetika tyagi <preetikatyagi@gmail.com>
>>> wrote:
>>>>
>>>> Hi All,
>>>>
>>>> I am little confused about the internal working of PostgreSQL. There is
>>>> a parameter shared_buffer in postgres.conf and I am assuming that it is used
>>>> for buffer management in PostgreSQL. If there is a need to bring in a new
>>>> page in the buffer and size exceeds the shared_buffer limit, a victim dirty
>>>> page will be written back to the disk.
>>>>
>>>> However, I have read on many links that PostgreSQL depends on the OS for
>>>> caching. (http://www.varlena.com/GeneralBits/Tidbits/perf.html#shbuf)
>>>>
>>>> So my question is, the actual limit of the shared buffer will be defined
>>>> by OS or the shared_buffer parameter in the postgres.conf to figure whether
>>>> a victim dirty page needs to be selected for disk write or not?
>>>>
>>>> Thanks!
>>>
>>
>
>



--
Cédric Villemain               2ndQuadrant
http://2ndQuadrant.fr/     PostgreSQL : Expertise, Formation et Support