Re: [GENERAL] Moving a large DB (> 500GB) to another DB with different locale

På onsdag 13. januar 2016 kl. 03:58:27, skrev John McKown < john.archie.mck...@gmail.com >: On Tue, Jan 12, 2016 at 4:09 PM, Andreas Joseph Krogh > wrote: På tirsdag 12. januar 2016 kl. 22:56:56, skrev John R

Re: [GENERAL] Function error

Sachin Srivastava wrote: > I am getting the below error for function, please see the bold line in > "Function code", please suggest > what I will do to correct this code. > > --- > > ERROR: syntax error at or near "*" > LINE 35: SELECT * from logError(msg text) is >

Re: [GENERAL] Code of Conduct: Is it time?

On Wed, Jan 13, 2016 at 9:00 AM, John R Pierce wrote: > On 1/12/2016 11:32 PM, Chris Travers wrote: > >> >> One of the nice things about the Ubuntu Code of Conduct is that it >> focuses primarily on the positive. It is long, perhaps overly verbose, but >> it does focus on

[GENERAL] Function error

Dear Team, I am getting the below error for function, please see the bold line in "Function code", please suggest what I will do to correct this code. --- ERROR: syntax error at or near "*" LINE 35: SELECT * from logError(msg text) is ^ CONTEXT: invalid

Re: [GENERAL] [BUGS] permissions.

[not a bug, moving to -general] On Tue, Jan 12, 2016 at 12:46 PM, Frik Brits wrote: > Hello, > What is to be done here, below between the lines is an excerpt from my > bash console if I want to start PostgreSQL. > > >

Re: [GENERAL] Code of Conduct: Is it time?

On 1/12/2016 11:32 PM, Chris Travers wrote: One of the nice things about the Ubuntu Code of Conduct is that it focuses primarily on the positive. It is long, perhaps overly verbose, but it does focus on what the community wants rather than what the community wants to avoid. It is easy to

Re: [GENERAL] Moving a large DB (> 500GB) to another DB with different locale

På onsdag 13. januar 2016 kl. 04:17:03, skrev Vick Khera >:   On Tue, Jan 12, 2016 at 4:20 PM, Andreas Joseph Krogh > wrote: I'm moving a > 500GB DB to another server which is initdb'ed with a different

Re: [GENERAL] WIP: CoC V5

This looks much better than earlier editions. I think this one I could avoid violating without too much trouble. When we talk about community commons, does that include physical spaces? I'm trying to visualize what "safe" means in a virtual space and I'm drawing a blank picture.

Re: [GENERAL] WIP: CoC V5

On 13 January 2016 at 03:10, Tom Lane wrote: > The "disparaging remarks" part of this could easily be taken to forbid > technical criticism of any sort, eg "this patch is bad because X,Y, and > Z", even when X,Y, and Z are perfectly neutral technical points. "Of any > kind"

Re: [GENERAL] WIP: CoC V5

On 13 January 2016 at 13:05, Regina Obe wrote: >> Perhaps you could add something about valuing contributions from and making >> allowances for those with less expertise. > > I agree it's hard to even talk about just technical without hurting someone's > feelings, I don't

Re: [GENERAL] WIP: CoC V5

On Wed, Jan 13, 2016 at 2:05 PM, Regina Obe wrote: > > > On 13 January 2016 at 03:10, Tom Lane wrote: > >> The "disparaging remarks" part of this could easily be taken to forbid > >> technical criticism of any sort, eg "this patch is bad because X,Y, > >> and

[GENERAL] Changing varchar length by manipulating pg_attribute

Hi I have a database in which I'd like to increase the length of a varchar column. Unfortunately, the column is used in various views which then are used in other views, so doing this with ALTER TABLE ALTER COLUMN TYPE is quite a lot of work. I have found this suggestion

Re: [GENERAL] WIP: CoC V5

> On 13 January 2016 at 03:10, Tom Lane wrote: >> The "disparaging remarks" part of this could easily be taken to forbid >> technical criticism of any sort, eg "this patch is bad because X,Y, >> and Z", even when X,Y, and Z are perfectly neutral technical points. >> "Of

Re: [GENERAL] WIP: CoC V5

On 13 January 2016 at 13:55, Chris Travers wrote: > People *can* take offense when you say their code is not good enough, > particularly when it is true, Well I would hope that you wouldn't say so if it wasn't. :) My point is that most people won't be posting code here

[GENERAL] [BDR] Best practice to automatically abort a DDL operation when one node is down

Hello all, I am using BDR with two nodes 1 and 2. If I issue a DDL operation in node 1 when node 2 is down, for example: CREATE TABLE test (i int PRIMARY KEY); (1) all other transactions fail with the following error: Database is locked against DDL operations The problem is that the (1)

[GENERAL] plpython3 package absent in 9.5 repository

I don't know the policy of package inclusion in the repositories. The plpython3 package exists in the Fedora repository: http://mirror.globo.com/fedora/linux/updates/22/x86_64/p/ But it does not exist in the Postgresql repository:

Re: [GENERAL] Moving a large DB (> 500GB) to another DB with different locale

On 1/13/16 2:03 AM, Andreas Joseph Krogh wrote: Seems like pglogical migth be better? http://2ndquadrant.com/en/resources/pglogical/ It's available for 9.4 also. It would certainly be faster. It's also less tested than Slony is though, as it's fairly new. If it was me, I'd use pg_logical. --

Re: [GENERAL] Changing varchar length by manipulating pg_attribute

On 1/13/16 5:59 AM, Christian Ramseyer wrote: UPDATE pg_attribute SET atttypmod = 35+4 -- new desired length + 4 WHERE attrelid = 'TABLE1'::regclass AND attname = 'COL1'; I don't know of any reason that wouldn't work. Note that you might have to make the same change to all the views too.

[GENERAL] pg_dump problem with dropped NOT NULL on child table

Please consider the following: - Debian - server is PG 9.4.5 - client (psql/pg_dump/libpq5) is 9.5 create table parent ( not_null_in_parent integer not null ); create table child() inherits (parent); alter table child alter column

Re: [GENERAL] Call postgres PL/Python stored function from another PL/Python block.

Ok, thank you very much. 2016-01-13 19:04 GMT+01:00 Adrian Klaver : > On 01/13/2016 07:47 AM, lodopidolo wrote: > >> Hello. It is possible to call al PL/Python stored function natively from >> another PL/Python function? >> >> Something like: >> >> |createorreplace

Re: [GENERAL] BDR install broken on Ubuntu 14.04

On 13-01-16 17:48, LOUBRADOU, Rémy wrote: Hi, I have the same issue. It was working fine a month ago. I wrote a cookbook with chef to install BDR a month ago since then I did not change the cookbook but now I got this error. This seem to be the heart of the issue: Removing obsolete

[GENERAL] Fwd: Data Packaging/Data Unpacking

Appologies, for posting it again, but I didn't get any responses so far. Looks like I posted it too late in the evening and it went not noticed. Oleg -- Forwarded message -- From: oleg yusim Date: Tue, Jan 12, 2016 at 10:00 PM Subject: Data Packaging/Data

Re: [GENERAL] Query Questions - PostgreSQL

Anyone who can help with this please? Thanks Sent from my phone > On 13/01/2016, at 3:30 PM, Saulo Merlo wrote: > > So.. I have a Query that is taking too long to complete. > > OLD QUERY: > >> SELECT > >> file.inode_idAS file_id, > >>

[GENERAL] Exclude bdr data from dump

Hello, I want to exclude all bdr data from a database dump. My command is: $ bdr_dump -Fp -h localhost -U postgres mydb -f /tmp/mydb.data --data-only --exclude-table='bdr* This results in bdr data being included. Including only bdr results in an error: $ bdr_dump -Fp -h localhost -U

Re: [GENERAL] Synchronous replication

On 01/13/2016 12:28 PM, jwienc...@comcast.net wrote: Is it possible to set up synchronous replication to more that one node in a cluster? Or, am I limited to one synchronous node and one asynchronous node? Yes you can have N number of synchronous slaves. However, be careful. It can be a huge

Re: [GENERAL] WIP: CoC V5

On Wed, 13 Jan 2016 09:15:45 -0500 "Regina Obe" wrote: > For example if Tom makes some snide remark like "Do all Bostonians > program this way?" Why not simply discuss the code, with no value judgment about the coder? "The strcpy() in the foo() function can cause intermittent

Re: [GENERAL] Data Packaging/Data Unpacking

On Wed, Jan 13, 2016 at 2:57 PM, oleg yusim wrote: > Say, I got network package. The package was decrypted by OpenSSL. Where this > data are, physically, at this moment? Process-local memory for the PostgreSQL backend process associated with the database connection. >

Re: [GENERAL] Blocked updates and background writer performance

On Wed, Jan 13, 2016 at 9:48 AM Vick Khera wrote: > That was my intuition too. Not enough I/O available from the hardware for > the workload requested. > > As recommended, log your checkpoints and try tuning them to spread the > load. > Thanks guys, will turn on checkpoint

Re: [GENERAL] New Slave - timeline ERROR

Hi guys.. I started a new PG_BASEBACKUP and it's working now.. The problem was the line: standby_mode = on on the recovery.conf on the STANDBY server. After the basebackup, I comented this line and started the postgreSQL. BUT, you shouldn't do that. On the last time I didn't comment and it

Re: [GENERAL] Unable to build python extension with PGXS

On 1/12/16 10:04 PM, Jim Nasby wrote: Attempting to build a python extension, I'm getting: Undefined symbols for architecture x86_64: "_PyErr_Clear", referenced from: _PLyNdarray_FromDatum in pg_ndarray.o _PLyObject_To_ndarray in pg_ndarray.o "_PyImport_ImportModule",

Re: [GENERAL] plv8 binaries for PostgreSQL 9.5 on Windows

On 01/13/2016 08:25 AM, Igal @ Lucee.org wrote: On 1/12/2016 12:57 PM, Adrian Klaver wrote: On 01/11/2016 05:50 PM, Igal @ Lucee.org wrote: Where can I find binaries for plv8 for PostgreSQL 9.5 for Windows? The only Windows binaries I could find where for 9.4, here:

Re: [GENERAL] Data Packaging/Data Unpacking

On Tue, Jan 12, 2016 at 10:00 PM, oleg yusim wrote: > Important: let's assume data at rest is encrypted using EFS and data at > transit is encrypted using ciphers, provided by OpenSSL. > > So, with that in mind, please, help me to understand movement and location > of the

Re: [GENERAL] pg_dump problem with dropped NOT NULL on child table

Karsten Hilbert writes: > Please consider the following: > create table parent ( > not_null_in_parent integer not null > ); > create table child() inherits (parent); > alter table child > alter column

Re: [GENERAL] Synchronous replication

On Thu, Jan 14, 2016 at 9:34 AM, Joshua D. Drake wrote: > On 01/13/2016 12:28 PM, jwienc...@comcast.net wrote: >> >> Is it possible to set up synchronous replication to more that one node >> in a cluster? Or, am I limited to one synchronous node and one >> asynchronous

Re: [GENERAL] Moving a large DB (> 500GB) to another DB with different locale

På onsdag 13. januar 2016 kl. 19:18:20, skrev Jim Nasby < jim.na...@bluetreble.com >: On 1/13/16 2:03 AM, Andreas Joseph Krogh wrote: > Seems like pglogical migth be better? > http://2ndquadrant.com/en/resources/pglogical/ > It's available for 9.4 also. It

Re: [GENERAL] Data Packaging/Data Unpacking

Kevin, Thank you for your reply. I understand what you are saying, but I guess I need a bit deeper understanding for my assessment. Let's dive a bit here: Say, I got network package. The package was decrypted by OpenSSL. Where this data are, physically, at this moment? Opposite situation: we

Re: [GENERAL] Function error

On 01/13/2016 01:18 AM, Sachin Srivastava wrote: > > Dear Team, > > I am getting the below error for function, please see the bold line in > "Function code", please suggest what I will do to correct this code. > > --- > > ERROR: syntax error at or near "*" > LINE 35:

Re: [GENERAL] pg_dump problem with dropped NOT NULL on child table

On 01/13/2016 11:38 AM, Karsten Hilbert wrote: Please consider the following: - Debian - server is PG 9.4.5 - client (psql/pg_dump/libpq5) is 9.5 create table parent ( not_null_in_parent integer not null ); create table child() inherits (parent);

[GENERAL] Synchronous replication

Is it possible to set up synchronous replication to more that one node in a cluster? Or, am I limited to one synchronous node and one asynchronous node? Assuming my replicated databases are on nodes: testqa1, testqa2. What would my synchronous_standby_names entry in my postgresql.conf on my

Re: [GENERAL] Function error

On Wed, Jan 13, 2016 at 10:06 PM, Sachin Srivastava wrote: > Dear Albe, > > Thanks for your support !!! > > In my function the problem is that global variables defined inside the > function. These variables are visible to functions defined inside a > function. > ​Holy

Re: [GENERAL] Data Packaging/Data Unpacking

On Wed, Jan 13, 2016 at 2:19 PM, Kevin Grittner wrote: > On Wed, Jan 13, 2016 at 2:57 PM, oleg yusim wrote: > > > Say, I got network package. The package was decrypted by OpenSSL. Where > this > > data are, physically, at this moment? > > Process-local

Re: [GENERAL] Data Packaging/Data Unpacking

OK, Kevin, David, Thanks you very much for explanation. Now who is the owner of this process? My understanding is, data then located physically in RAM, in the memory stack assigned by OS to this process. Now the question is who owns the process? Thanks, Oleg On Wed, Jan 13, 2016 at 3:29 PM,

Re: [GENERAL] Function error

Dear Albe, Thanks for your support !!! In my function the problem is that global variables defined inside the function. These variables are visible to functions defined inside a function. If we move these inner functions to outside of the main function, they will lose the visibility of the

Re: [GENERAL] WIP: CoC V5

Steve Litt writes: > To me it's simple... > Disallow "You " > Disallow "Your code " > Encourage "It would be better if your code because > ." That's fine as practice, but I don't think we want to get anywhere near being that detailed in the CoC per se. If we start

Re: [GENERAL] WIP: CoC V5

On 01/13/2016 01:27 PM, Tom Lane wrote: Steve Litt writes: To me it's simple... Disallow "You " Disallow "Your code " Encourage "It would be better if your code because ." That's fine as practice, but I don't think we want to get anywhere near being that detailed

Re: [GENERAL] Data Packaging/Data Unpacking

On Wed, Jan 13, 2016 at 3:54 PM, oleg yusim wrote: > Answer "postgres" would suffice. But the user would not always be "postgres". To be accurate, it is the user which owns the files for the "cluster" (database instance) and which runs the database service. If a machine

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

On 01/13/2016 02:24 PM, Tom Lane wrote: "Williamson, Michael" writes: I'm attempting to drop a trigger that may or may not exist, so am using the "IF EXISTS" clause. Â This works fine for tables, views, functions, domains, and types, but for some reason seems to

Re: [GENERAL] Data Packaging/Data Unpacking

On Wed, Jan 13, 2016 at 2:37 PM, oleg yusim wrote: > OK, Kevin, David, > > Thanks you very much for explanation. Now who is the owner of this > process? My understanding is, data then located physically in RAM, in the > memory stack assigned by OS to this process. Now the

Re: [GENERAL] Moving a large DB (> 500GB) to another DB with different locale

On 1/13/16 2:39 PM, Andreas Joseph Krogh wrote: Where can I find more info about how to use and configure pg_logical to replicate a 9.4 DB to 9.5? http://2ndquadrant.com/en/resources/pglogical/ -- Jim Nasby, Data Architect, Blue Treble Consulting, Austin TX Experts in Analytics, Data

Re: [GENERAL] Code of Conduct: Is it time?

On Tue, Jan 5, 2016 at 08:47:16AM -0800, Joshua Drake wrote: > Hello, > > I had a hard time writing this email. I think Code of Conducts are > non-essential, a waste of respectful people's time and frankly if > you are going to be a jerk, our community will call you out on it. > Unfortunately a

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

On 01/13/2016 03:25 PM, Williamson, Michael wrote: Ccing list Is that happening on all triggers you are trying to DROP, or just this one? All, consistently. If just this one, you might want to check to see if the trigger was created with a quoted mixed case name and therefore needs to have

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

"Williamson, Michael" writes: > I'm attempting to drop a trigger that may or may not exist, so am using > the "IF EXISTS" clause.  This works fine for tables, views, functions, > domains, and types, but for some reason seems to be ignored for > triggers.  I'd

Re: [GENERAL] plpython3 package absent in 9.5 repository

Hi Clodoaldo, It is a shame that I totally skipped that. Thanks for the report. I am working on the patch now. I will commit it as soon as it is ready. Regards, Devrim  On Wed, 2016-01-13 at 10:44 -0200, Clodoaldo Neto wrote: > I don't know the policy of package inclusion in the

[GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

I'm attempting to drop a trigger that may or may not exist, so am using the "IF EXISTS" clause.  This works fine for tables, views, functions, domains, and types, but for some reason seems to be ignored for triggers.  I'd expect to see more about this online if it were a bug, so I'm thinking I may

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

On 01/13/2016 02:14 PM, Williamson, Michael wrote: I'm attempting to drop a trigger that may or may not exist, so am using the "IF EXISTS" clause. This works fine for tables, views, functions, domains, and types, but for some reason seems to be ignored for triggers. I'd expect to see more

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

On 14 January 2016 at 11:32, Adrian Klaver wrote: > On 01/13/2016 02:24 PM, Tom Lane wrote: > >> "Williamson, Michael" writes: >> >>> I'm attempting to drop a trigger that may or may not exist, so am using >>> the "IF EXISTS" clause. Â

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

On 01/13/2016 02:51 PM, David Rowley wrote: On 14 January 2016 at 11:32, Adrian Klaver > wrote: On 01/13/2016 02:24 PM, Tom Lane wrote: "Williamson, Michael"

Re: [GENERAL] Synchronous replication

Thank you all for the quick replies. John Wiencek On 1/13/16, 2:43 PM, "Thomas Munro" wrote: >On Thu, Jan 14, 2016 at 9:34 AM, Joshua D. Drake >wrote: >> On 01/13/2016 12:28 PM, jwienc...@comcast.net wrote: >>> >>> Is it possible to set

Re: [GENERAL] Data Packaging/Data Unpacking

David, Answer "postgres" would suffice. I have fairly decent idea of what happens below that (you are right system utility - Memory Manager is what marshaling data back and forth to RAM, abstracting absolute addresses from application along the way, and once in RAM, security kernel of the system

Re: [GENERAL] Data Packaging/Data Unpacking

Kevin, It is understood. But doesn't change things from my perspective. It would anyway be user, with sole access, nobody can elevate privileges to and only root can su to. The name, in that case, is irrelevant, but answer postgres gave me idea of the protection level here. Thanks, Oleg On

Re: [GENERAL] v9.1, DROP TRIGGER IF EXISTS behaving oddly

On 14 January 2016 at 12:08, Adrian Klaver wrote: > On 01/13/2016 02:51 PM, David Rowley wrote: > >> On 14 January 2016 at 11:32, Adrian Klaver > > wrote: >> >> On 01/13/2016 02:24 PM, Tom Lane wrote: >>

Re: [GENERAL] WIP: CoC V5

On Tue, Jan 12, 2016 at 10:04 PM, Joshua D. Drake wrote: > On 01/12/2016 07:10 PM, Tom Lane wrote: >> Kevin Grittner writes: >>> * To maintain a safe, respectful, productive and collaborative >>> environment all participants must ensure that their

[GENERAL] Call postgres PL/Python stored function from another PL/Python block.

Hello. It is possible to call al PL/Python stored function natively from another PL/Python function? Something like: create or replace function f1() returns text as $$ return "hello"$$ language 'plpython3u'; do $$begin ... t = f1() ...end;$$ language 'plpython3u'; Regards.

Re: [GENERAL] Why PG uses nested-loop join when no indexes are available?

David Grelaud writes: > Statistics are not propagated when Common Table Expressions (CTE) are used. > The cardinality of a CTE is equal to 1 most of the time Really? The rest of this seems to be proceeding from completely false assumptions.

Re: [GENERAL] WIP: CoC V5

Kevin Grittner writes: > I still feel it is more productive to discuss a proposed document > than proposed language for some "motion to adopt". Agreed. We're trying to write a document, not a document about a document. > ... It applies > to all "collaborative space", which

Re: [GENERAL] WIP: CoC V5

> On 13 January 2016 at 03:10, Tom Lane > wrote: >> The "disparaging remarks" part of this could easily be taken to forbid >> technical criticism of any sort, eg "this patch is bad because X,Y, >> and Z", even when X,Y, and Z are perfectly neutral

Re: [GENERAL] WIP: CoC V5

Another anecdotal thing. Personal attacks sometimes soften the blow. Take that as you will. For example if Tom makes some snide remark like "Do all Bostonians program this way?" It would lessen the blow of the criticism of the code as I would think he's making fun of Bostonians coding

[GENERAL] Why PG uses nested-loop join when no indexes are available?

Hi, Statistics are not propagated when Common Table Expressions (CTE) are used. The cardinality of a CTE is equal to 1 most of the time so every joins with previously computed CTEs are done with the nested-loop algorithm. This seems to be really a risky choice, even without CTEs, according to

Re: [GENERAL] WIP: CoC V5

On 13 January 2016 at 14:15, Regina Obe wrote: > He can also make fun of my tabbing style and say "What's wrong with your > editor? Perhaps you need to use a different one or change the settings" You're right, what we've really been missing all these years is an editor flamewar

Re: [GENERAL] WIP: CoC V5

On Wed, Jan 13, 2016 at 3:15 PM, Regina Obe wrote: > Another anecdotal thing. Personal attacks sometimes soften the blow. Take > that as you will. > > > > For example if Tom makes some snide remark like "Do all Bostonians program > this way?" > > > > It would lessen the blow of

Re: [GENERAL] Why PG uses nested-loop join when no indexes are available?

Thank you for your fast response! I'm sorry, my vocabulary may be not correct and my "french approach" to explain my problem is not efficient ;). But the underestimation problem in complex queries is still there? ... and generates potential "dangerous" plans (nested loop). We thought at the

Re: [GENERAL] plv8 binaries for PostgreSQL 9.5 on Windows

On 1/12/2016 12:57 PM, Adrian Klaver wrote: On 01/11/2016 05:50 PM, Igal @ Lucee.org wrote: Where can I find binaries for plv8 for PostgreSQL 9.5 for Windows? The only Windows binaries I could find where for 9.4, here:

Re: [GENERAL] Call postgres PL/Python stored function from another PL/Python block.

Hi 2016-01-13 16:47 GMT+01:00 lodopidolo : > Hello. It is possible to call al PL/Python stored function natively from > another PL/Python function? > The function created by CREATE OR REPLACE FUNCTION statement can be called only via special interface, or via SQL

Re: [GENERAL] Call postgres PL/Python stored function from another PL/Python block.

On 1/13/16 9:47 AM, lodopidolo wrote: Hello. It is possible to call al PL/Python stored function natively from another PL/Python function? Stackoverflow is stupid and won't let me post there, but here's what you want: There is no special capability to call other plpython functions. You

Re: [GENERAL] WIP: CoC V5

On 01/13/2016 08:05 AM, Tom Lane wrote: My general reaction to the rest of this is that it's got the right idea, but it could be cut to about half the length and be better off for that. Short and sweet is the way, IMO. +1 Less words, more point. -- Command Prompt, Inc. -

Re: [GENERAL] Call postgres PL/Python stored function from another PL/Python block.

On 01/13/2016 07:47 AM, lodopidolo wrote: Hello. It is possible to call al PL/Python stored function natively from another PL/Python function? Something like: |createorreplace functionf1()returns text as$$return"hello"$$language 'plpython3u';| |do $$begin...t =f1()...end;$$language

Re: [GENERAL] Moving a large DB (> 500GB) to another DB with different locale

On Wed, Jan 13, 2016 at 3:03 AM, Andreas Joseph Krogh wrote: > Seems like pglogical migth be better? > http://2ndquadrant.com/en/resources/pglogical/ > > I would have no idea. I never used it, nor do I run RHEL (or any linux for that matter) which seems to be the only

Re: [GENERAL] Blocked updates and background writer performance

On Wed, Jan 13, 2016 at 11:51 AM, Jeff Janes wrote: > Anyway, it looks to me like you have a checkpoint problem. The checkpoint > overwhelms your IO system. The overwhelmed IO system then backs up into > the bgwriter. What you see in the bgwriter is just a symptom, not