Thread: 12 hour table vacuums
We vacuum only a few of our tables nightly, this one is the last one because it takes longer to run. I'll probably re-index it soon, but I would appreciate any advice on how to speed up the vacuum process (and the db in general). Okay, here's our system: postgres 8.1.4 Linux version 2.4.21 Red Hat Linux 3.2.3 8 GB ram Intel(R) Xeon(TM) CPU 3.20GHz Raid 5 autovacuum=off serves as the application server and database server server is co-located in another city, hardware upgrade is not currently an option Here's the table information: The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. It is probably our 'key' table in the database and gets called by almost every query (usually joined to others). The table gets updated only about 10 times a day. We were running autovacuum but it interfered with the updates to we shut it off. We vacuum this table nightly, and it currently takes about 12 hours to vacuum it. Not much else is running during this period, nothing that should affect the table. Here are the current non-default postgresql.conf settings: max_connections = 100 shared_buffers = 50000 work_mem = 9192 maintenance_work_mem = 786432 max_fsm_pages = 70000 vacuum_cost_delay = 200 vacuum_cost_limit = 100 bgwriter_delay = 10000 fsync = on checkpoint_segments = 64 checkpoint_timeout = 1800 effective_cache_size = 270000 random_page_cost = 2 log_destination = 'stderr' redirect_stderr = on client_min_messages = warning log_min_messages = warning stats_start_collector = off stats_command_string = on stats_block_level = on stats_row_level = on autovacuum = off autovacuum_vacuum_threshold = 2000 deadlock_timeout = 10000 max_locks_per_transaction = 640 add_missing_from = on As I mentioned, any insights into changing the configuration to optimize performance are most welcome. Thanks Ron
In response to Ron St-Pierre <ron.pgsql@shaw.ca>: > We vacuum only a few of our tables nightly, this one is the last one > because it takes longer to run. I'll probably re-index it soon, but I > would appreciate any advice on how to speed up the vacuum process (and > the db in general). I doubt anyone can provide meaningful advice without the output of vacuum verbose. > > Okay, here's our system: > postgres 8.1.4 > Linux version 2.4.21 > Red Hat Linux 3.2.3 > 8 GB ram > Intel(R) Xeon(TM) CPU 3.20GHz > Raid 5 > autovacuum=off > serves as the application server and database server > server is co-located in another city, hardware upgrade is not > currently an option > > Here's the table information: > The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. It > is probably our 'key' table in the database and gets called by almost > every query (usually joined to others). The table gets updated only > about 10 times a day. We were running autovacuum but it interfered with > the updates to we shut it off. We vacuum this table nightly, and it > currently takes about 12 hours to vacuum it. Not much else is running > during this period, nothing that should affect the table. > > Here are the current non-default postgresql.conf settings: > max_connections = 100 > shared_buffers = 50000 > work_mem = 9192 > maintenance_work_mem = 786432 > max_fsm_pages = 70000 > vacuum_cost_delay = 200 > vacuum_cost_limit = 100 > bgwriter_delay = 10000 > fsync = on > checkpoint_segments = 64 > checkpoint_timeout = 1800 > effective_cache_size = 270000 > random_page_cost = 2 > log_destination = 'stderr' > redirect_stderr = on > client_min_messages = warning > log_min_messages = warning > stats_start_collector = off > stats_command_string = on > stats_block_level = on > stats_row_level = on > autovacuum = off > autovacuum_vacuum_threshold = 2000 > deadlock_timeout = 10000 > max_locks_per_transaction = 640 > add_missing_from = on > > As I mentioned, any insights into changing the configuration to optimize > performance are most welcome. > > Thanks > > Ron > > ---------------------------(end of broadcast)--------------------------- > TIP 5: don't forget to increase your free space map settings > > > > > > -- Bill Moran Collaborative Fusion Inc. http://people.collaborativefusion.com/~wmoran/ wmoran@collaborativefusion.com Phone: 412-422-3463x4023 **************************************************************** IMPORTANT: This message contains confidential information and is intended only for the individual named. If the reader of this message is not an intended recipient (or the individual responsible for the delivery of this message to an intended recipient), please be advised that any re-use, dissemination, distribution or copying of this message is prohibited. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission. ****************************************************************
Ron St-Pierre <ron.pgsql@shaw.ca> writes:
> The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. It
> is probably our 'key' table in the database and gets called by almost
> every query (usually joined to others). The table gets updated only
> about 10 times a day. We were running autovacuum but it interfered with
> the updates to we shut it off. We vacuum this table nightly, and it
> currently takes about 12 hours to vacuum it. Not much else is running
> during this period, nothing that should affect the table.
Here is your problem:
> vacuum_cost_delay = 200
If you are only vacuuming when nothing else is happening, you shouldn't
be using vacuum_cost_delay at all: set it to 0. In any case this value
is probably much too high. I would imagine that if you watch the
machine while the vacuum is running you'll find both CPU and I/O load
near zero ... which is nice, unless you would like the vacuum to finish
sooner.
In unrelated comments:
> maintenance_work_mem = 786432
That seems awfully high, too.
> max_fsm_pages = 70000
And this possibly too low --- are you sure you are not leaking disk
space?
> stats_start_collector = off
> stats_command_string = on
> stats_block_level = on
> stats_row_level = on
These are not self-consistent.
regards, tom lane
Ron St-Pierre wrote: > Okay, here's our system: > postgres 8.1.4 Upgrade to 8.1.10 > Here's the table information: > The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. 60 indexes? You gotta be kidding. You really have 60 columns on which to scan? > vacuum_cost_delay = 200 > vacuum_cost_limit = 100 Isn't this a bit high? What happens if you cut the delay to, say, 10? (considering you've lowered the limit to half the default) -- Alvaro Herrera Developer, http://www.PostgreSQL.org/ "Someone said that it is at least an order of magnitude more work to do production software than a prototype. I think he is wrong by at least an order of magnitude." (Brian Kernighan)
On Tue, 2007-10-23 at 08:53 -0700, Ron St-Pierre wrote: > [snip] We were running autovacuum but it interfered with > the updates to we shut it off. This is not directly related to your question, but it might be good for your DB: you don't need to turn off autovacuum, you can exclude tables individually from being autovacuumed by inserting the appropriate rows in pg_autovacuum. See: http://www.postgresql.org/docs/8.1/static/catalog-pg-autovacuum.html We also do have here a few big tables which we don't want autovacuum to touch, so we disable them via pg_autovacuum. There are a few really big ones which change rarely - those we only vacuum via a DB wide vacuum in the weekend (which for us is a low activity period). If you say your table is only changed rarely, you might be OK too with such a setup... Cheers, Csaba.
Bill Moran wrote: > In response to Ron St-Pierre <ron.pgsql@shaw.ca>: > > >> We vacuum only a few of our tables nightly, this one is the last one >> because it takes longer to run. I'll probably re-index it soon, but I >> would appreciate any advice on how to speed up the vacuum process (and >> the db in general). >> > > I doubt anyone can provide meaningful advice without the output of > vacuum verbose. > > The cron job is still running /usr/local/pgsql/bin/vacuumdb -d imperial -t stock.fdata -v -z > /usr/local/pgsql/bin/fdata.txt I'll post the output when it's finished. Ron
"Ron St-Pierre" <ron.pgsql@shaw.ca> writes: > We vacuum only a few of our tables nightly, this one is the last one because it > takes longer to run. I'll probably re-index it soon, but I would appreciate any > advice on how to speed up the vacuum process (and the db in general). ... > vacuum_cost_delay = 200 Well speeding up vacuum isn't really useful in itself. In fact you have vacuum configured to run quite slowly by having vacuum_cost_delay set so high. You have it set to sleep 200ms every few pages. If you lower that it'll run faster but take more bandwidth away from the foreground tasks. > Here's the table information: > The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. For what it's worth NUMERIC columns take more space than you might expect. Figure a minimum of 12 bytes your rows are at about 1.5k each even if the non-numeric columns aren't large themselves. What are the other columns? > We were running autovacuum but it interfered with the updates to we shut it > off. Was it just the I/O bandwidth? I'm surprised as your vacuum_cost_delay is quite high. Manual vacuum doesn't do anything differently from autovacuum, neither should interfere directly with updates except by taking away I/O bandwidth. > We vacuum this table nightly, and it currently takes about 12 hours to > vacuum it. Not much else is running during this period, nothing that should > affect the table. Is this time increasing over time? If once a day isn't enough then you may be accumulating more and more dead space over time. In which case you may be better off running it during prime time with a large vacuum_cost_delay (like the 200 you have configured) rather than trying to get to run fast enough to fit in the off-peak period. > deadlock_timeout = 10000 I would not suggest having this quite this high. Raising it from the default is fine but having a value larger than your patience is likely to give you the false impression that something is hung if you should ever get a deadlock. -- Gregory Stark EnterpriseDB http://www.enterprisedb.com
In response to Ron St-Pierre <ron.pgsql@shaw.ca>: > Bill Moran wrote: > > In response to Ron St-Pierre <ron.pgsql@shaw.ca>: > > > > > >> We vacuum only a few of our tables nightly, this one is the last one > >> because it takes longer to run. I'll probably re-index it soon, but I > >> would appreciate any advice on how to speed up the vacuum process (and > >> the db in general). > >> > > > > I doubt anyone can provide meaningful advice without the output of > > vacuum verbose. Understood, however I may have spoken too soon. It appears that Tom found an obvious issue with your config that seems likely to be the problem. -- Bill Moran Collaborative Fusion Inc. http://people.collaborativefusion.com/~wmoran/ wmoran@collaborativefusion.com Phone: 412-422-3463x4023
Tom Lane wrote: > Here is your problem: > > >> vacuum_cost_delay = 200 >> > > If you are only vacuuming when nothing else is happening, you shouldn't > be using vacuum_cost_delay at all: set it to 0. In any case this value > is probably much too high. I would imagine that if you watch the > machine while the vacuum is running you'll find both CPU and I/O load > near zero ... which is nice, unless you would like the vacuum to finish > sooner. > Yeah, I've noticed that CPU, mem and I/O load are really low when this is running. I'll change that setting. > In unrelated comments: > > >> maintenance_work_mem = 786432 >> > > That seems awfully high, too. > > Any thoughts on a more reasonable value? >> max_fsm_pages = 70000 >> > > And this possibly too low --- The default appears to be 20000, so I upped it to 70000. I'll try 160000 (max_fsm_relations*16). > are you sure you are not leaking disk > space? > > What do you mean leaking disk space? >> stats_start_collector = off >> stats_command_string = on >> stats_block_level = on >> stats_row_level = on >> > > These are not self-consistent. > > regards, tom lane > >
Gregory Stark wrote: > "Ron St-Pierre" <ron.pgsql@shaw.ca> writes: > > >> We vacuum only a few of our tables nightly, this one is the last one because it >> takes longer to run. I'll probably re-index it soon, but I would appreciate any >> advice on how to speed up the vacuum process (and the db in general). >> > ... > >> vacuum_cost_delay = 200 >> > > Well speeding up vacuum isn't really useful in itself. In fact you have vacuum > configured to run quite slowly by having vacuum_cost_delay set so high. You > have it set to sleep 200ms every few pages. If you lower that it'll run faster > but take more bandwidth away from the foreground tasks. > It's okay if it uses a lot of resources, because it's scheduled to run during the night (our slow time). Because most of the important queries running during the day use this table, I want the vacuum analzye finished ASAP. > >> Here's the table information: >> The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. >> > > For what it's worth NUMERIC columns take more space than you might expect. > Figure a minimum of 12 bytes your rows are at about 1.5k each even if the > non-numeric columns aren't large themselves. What are the other columns? > The NUMERIC columns hold currency related values, with values ranging from a few cents to the billions, as well as a few negative numbers. > >> We were running autovacuum but it interfered with the updates to we shut it >> off. >> > > Was it just the I/O bandwidth? I'm surprised as your vacuum_cost_delay is > quite high. Manual vacuum doesn't do anything differently from autovacuum, > neither should interfere directly with updates except by taking away > I/O bandwidth. > > I don't know what the problem was. I tried to exclude certain tables from autovacuuming, but it autovacuumed anyway. >> We vacuum this table nightly, and it currently takes about 12 hours to >> vacuum it. Not much else is running during this period, nothing that should >> affect the table. >> > > Is this time increasing over time? If once a day isn't enough then you may be > accumulating more and more dead space over time. In which case you may be > better off running it during prime time with a large vacuum_cost_delay (like > the 200 you have configured) rather than trying to get to run fast enough to > fit in the off-peak period. > > >> deadlock_timeout = 10000 >> > > I would not suggest having this quite this high. Raising it from the default > is fine but having a value larger than your patience is likely to give you the > false impression that something is hung if you should ever get a deadlock. > > Good point. I'll look into this. Thanks Ron
Alvaro Herrera wrote: > Ron St-Pierre wrote: > > >> Okay, here's our system: >> postgres 8.1.4 >> > > Upgrade to 8.1.10 > Any particular fixes in 8.1.10 that would help with this? > >> Here's the table information: >> The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. >> > > 60 indexes? You gotta be kidding. You really have 60 columns on which > to scan? > > Really. 60 indexes. They're the most commonly requested columns for company information (we believe). Any ideas on testing our assumptions about that? I would like to know definitively what are the most popular columns. Do you think that rules would be a good approach for this? (Sorry if I'm getting way off topic here) >> vacuum_cost_delay = 200 >> vacuum_cost_limit = 100 >> > > Isn't this a bit high? What happens if you cut the delay to, say, 10? > (considering you've lowered the limit to half the default) > > Yes, Tom pointed this out too. I'll lower it and check out the results. Ron
On Tue, 23 Oct 2007 10:00:05 -0700
Ron St-Pierre <ron.pgsql@shaw.ca> wrote:
> Alvaro Herrera wrote:
> > Ron St-Pierre wrote:
> >
> >
> >> Okay, here's our system:
> >> postgres 8.1.4
> >>
> >
> > Upgrade to 8.1.10
> >
> Any particular fixes in 8.1.10 that would help with this?
> >
> >> Here's the table information:
> >> The table has 140,000 rows, 130 columns (mostly NUMERIC), 60
> >> indexes.
> >
> > 60 indexes? You gotta be kidding. You really have 60 columns on
> > which to scan?
> >
> >
> Really. 60 indexes. They're the most commonly requested columns for
> company information (we believe). Any ideas on testing our
> assumptions about that? I would like to know definitively what are
> the most popular columns. Do you think that rules would be a good
> approach for this? (Sorry if I'm getting way off topic here)
I suggest you:
1. Turn on stats and start looking in the stats columns to see what
indexes are actually being used.
2. Strongly review your normalization :)
Joshua D. Drake
--
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 24x7/Emergency: +1.800.492.2240
PostgreSQL solutions since 1997 http://www.commandprompt.com/
UNIQUE NOT NULL
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
Attachment
On Tue, 2007-10-23 at 08:53 -0700, Ron St-Pierre wrote: > The table gets updated only > about 10 times a day. So why are you VACUUMing it nightly? You should do this at the weekend every 3 months... 8.1 is slower at VACUUMing indexes than later releases, so 60 indexes are going to hurt quite a lot. The default maintenance_work_mem is sufficient for this table. -- Simon Riggs 2ndQuadrant http://www.2ndQuadrant.com
Ron St-Pierre wrote: > Gregory Stark wrote: >>> We were running autovacuum but it interfered with the updates to we >>> shut it off. >> >> Was it just the I/O bandwidth? I'm surprised as your >> vacuum_cost_delay is quite high. Manual vacuum doesn't do anything >> differently from autovacuum, neither should interfere directly with >> updates except by taking away I/O bandwidth. >> > I don't know what the problem was. I tried to exclude certain tables > from autovacuuming, but it autovacuumed anyway. Probably because of Xid wraparound issues. Now that you're vacuuming weekly it shouldn't be a problem. (It's also much less of a problem in 8.2). -- Alvaro Herrera http://www.CommandPrompt.com/ The PostgreSQL Company - Command Prompt, Inc.
Ron St-Pierre wrote: > Alvaro Herrera wrote: >> Ron St-Pierre wrote: >> >> >>> Okay, here's our system: >>> postgres 8.1.4 >>> >> >> Upgrade to 8.1.10 >> > Any particular fixes in 8.1.10 that would help with this? I don't think so, but my guess is that you really want to avoid the autovacuum bug which makes it vacuum without FREEZE on template0, that has caused so many problems all over the planet. >>> Here's the table information: >>> The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. >> >> 60 indexes? You gotta be kidding. You really have 60 columns on which >> to scan? >> >> > Really. 60 indexes. They're the most commonly requested columns for company > information (we believe). Any ideas on testing our assumptions about that? > I would like to know definitively what are the most popular columns. Do you > think that rules would be a good approach for this? (Sorry if I'm getting > way off topic here) As Josh Drake already said, you can check pg_stat* views to see which indexes are not used. Hard to say anything else without seeing the definition. -- Alvaro Herrera http://www.CommandPrompt.com/ PostgreSQL Replication, Consulting, Custom Development, 24x7 support
In article <471E26E4.7040804@shaw.ca>, Ron St-Pierre <ron.pgsql@shaw.ca> writes: >> For what it's worth NUMERIC columns take more space than you might expect. >> Figure a minimum of 12 bytes your rows are at about 1.5k each even if the >> non-numeric columns aren't large themselves. What are the other columns? > The NUMERIC columns hold currency related values, with values ranging > from a few cents to the billions, as well as a few negative numbers. What's the required precision? If it's just cents (or maybe tenths thereof), you could use BIGINT to store the amount in this precision. This would give you exact values with much less space.
Ron St-Pierre wrote: > We vacuum only a few of our tables nightly, this one is the last one > because it takes longer to run. I'll probably re-index it soon, but I > would appreciate any advice on how to speed up the vacuum process (and > the db in general). I am a novice to postgreSQL, so I have no answers for you. But for my own education, I am confused by some of your post. > > Okay, here's our system: > postgres 8.1.4 I have postgresql-8.1.9-1.el5 > Linux version 2.4.21 I imagine you mean Linux kernel version; I have 2.6.18-8.1.15.el5PAE > Red Hat Linux 3.2.3 I have no clue what this means. Red Hat Linux 3 must have been in the early 1990s. RHL 5 came out about 1998 IIRC. Red Hat Enterprise Linux 3, on the other hand, was not numbered like that, as I recall. I no longer run that, but my current RHEL5 is named like this: Red Hat Enterprise Linux Server release 5 (Tikanga) and for my CentOS 4 system, it is CentOS release 4.5 (Final) Did RHEL3 go with the second dot in their release numbers? I do not remember that. > 8 GB ram > Intel(R) Xeon(TM) CPU 3.20GHz > Raid 5 > autovacuum=off Why would you not have that on? > serves as the application server and database server > server is co-located in another city, hardware upgrade is not > currently an option > > Here's the table information: > The table has 140,000 rows, 130 columns (mostly NUMERIC), 60 indexes. I have designed databases, infrequently, but since the late 1970s. In my experience, my tables had relatively few columns, rarely over 10. Are you sure this table needs so many? Why not, e.g., 13 tables averaging 10 columns each? OTOH, 140,000 rows is not all that many. I have a 6,000,000 row table in my little database on my desktop, and I do not even consider that large. Imagine the size of a database belonging to the IRS, for example. Surely it would have at least one row for each taxpayer and each employer (possibly in two tables, or two databases). Here are the last few lines of a VACUUM VERBOSE; command for that little database. The 6,000,000 row table is not in the database at the moment, nor are some of the other tables, but two relatively (for me) large tables are. CPU 0.00s/0.00u sec elapsed 0.00 sec. INFO: free space map contains 166 pages in 76 relations DETAIL: A total of 1280 page slots are in use (including overhead). 1280 page slots are required to track all free space. Current limits are: 40000 page slots, 1000 relations, using 299 KB. VACUUM stock=> select count(*) from ranks; [table has 10 columns] count -------- 981030 (1 row) stock=> select count(*) from ibd; [table has 8 columns] count --------- 1099789 (1 row) And this is the time for running that psql process, most of which was consumed by slow typing on my part. real 1m40.206s user 0m0.027s sys 0m0.019s My non-default settings for this are # - Memory - shared_buffers = 251000 work_mem = 32768 max_fsm_pages = 40000 I have 8GBytes RAM on this machine, and postgreSQL is the biggest memory user. I set shared_buffers high to try to get some entire (small) tables in RAM and to be sure there is room for indices. -- .~. Jean-David Beyer Registered Linux User 85642. /V\ PGP-Key: 9A2FC99A Registered Machine 241939. /( )\ Shrewsbury, New Jersey http://counter.li.org ^^-^^ 08:40:01 up 1 day, 58 min, 1 user, load average: 4.08, 4.13, 4.17