I am having a hang condition every time I try to retrieve a large
records (bytea) data from a table
The OS is a 5.11 snv_134 i86pc i386 i86pc Solaris with 4GB memory
running Postgresql 8.4.3 with a standard postgresql.conf file (nothing
has been changed)
I have the following table called doc_table
Column | Type | Modifiers |
Storage | Description
------------------------+--------------------------------+---------------------------------------
id | numeric | not null | main |
file_n | character varying(4000) | |
extended |
create_date | timestamp(6) without time zone | not null
default (clock_timestamp())
::timestamp(0)without time zone | plain |
desc | character varying(4000) | |
extended |
doc_cc | character varying(120) | not null |
extended |
by | numeric | not null | main |
doc_data | bytea | |
extended |
mime_type_id | character varying(16) | not null |
extended |
doc_src | text | |
extended |
doc_stat | character varying(512) | not null
default 'ACTIVE'::character varying |
extended |
Indexes:
"documents_pk" PRIMARY KEY, btree (document_id)
A while ago the some developers inserted several records with a
document (stored in doc_Data) that was around 400 - 450 MB each. Now
when you do a select * (all) from this table you get a hang and the
system becomes unresponsive. Prior to these inserts, a select * (all,
no where clause) worked. I'm also told a select * from doc_table
where id = xxx still works. I haven't seen any error message in the
postgresql log files.
So I'm not sure how to find these bad records and why I am getting a
hang. Since this postgresql is running with the default config files
could I be running out of a resource? If so I'm not sure how to or
how much to add to these resources to fix this problem since I have
very little memory on this system. Does anyone have any ideas why I
am getting a hang. Thanks