Cannot read block error.

Hello,

When was the last time you ran a reindex? Or a vacuum / vacuum full?

Sincerely,

Joshua D. Drake

On Sat, 14 Feb 2004, Jason Essington wrote:

I am running PostgreSQL 7.3.3 on OS X Server 10.2

The database has been running just fine for quite some time now, but
this morning it began pitching the error:
ERROR: cannot read block 176 of tfxtrade_details: Numerical result
out of range
any time the table tfxtrade_details is accessed.

A description of the table is at the end of this email

I have a backup from last night, so I haven't lost much data (if any),
but I am curious if there is a way to recover from this (beyond
restoring from backup) and how I would go about figuring out what
caused it to prevent it from happening again.

I will keep a copy of the data directory if anyone wants me to do any
analysis on it (I will need instructions).

Any insights would be appreciated.

Thanks

Jason Essington
jaessing@greenrivercomputing.com

hedgehog=# \d tfxtrade_details
Table "public.tfxtrade_details"
Column | Type | Modifiers
---------------+--------------------------+-----------
rid | integer | not null
clientid | integer |
tradeid | integer |
rollid | integer |
rollpct | numeric(10,8) |
expdetailid | integer |
expid | integer |
contractpct | numeric(10,8) |
contractamt | numeric(18,2) |
origpct | numeric(10,8) |
origamt | numeric(18,2) |
acctgperiod | integer |
acctgperiodid | integer |
editdate | timestamp with time zone |
edituserid | character varying(48) |
parentid | integer |
entityid | integer |
tradedate | date |
maturitydate | date |
strategyid | integer |
currencyid | integer |
Indexes: tfxtrade_details_pkey primary key btree (rid),
tfxlinks_entityid_index btree (entityid),
tfxlinks_expdetailid_index btree (expdetailid),
tfxlinks_expid_index btree (expid),
tfxlinks_mdate_index btree (maturitydate),
tfxlinks_parentid_index btree (parentid),
tfxlinks_strategy_index btree (strategyid),
tfxlinks_tradeid_index btree (tradeid)
Triggers: RI_ConstraintTrigger_30891,
RI_ConstraintTrigger_30894,
tfxdetail_delete_trigger

---------------------------(end of broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faqs/FAQ.html

--
Co-Founder
Command Prompt, Inc.
The wheel's spinning but the hamster's dead

Both vacuum [full] and reindex fail with that same error.

vacuum is run regularly via a cron job.

-jason
On Feb 14, 2004, at 2:29 PM, Joshua D. Drake wrote:

Hello,

There are a couple of things it could be. I would suggest that you take
down the database, start it up with -P? (I think it is -o '-P' it might
be -p '-O' I don't recall) and try and reindex the database itself.

You can also do a vacuuum verbose and see if you get some more errors you
may have a corrupt system index that needs to be reindexed.

Sincerely,

Johsua D. Drake

On Sat, 14 Feb 2004, Jason Essington wrote:

Both vacuum [full] and reindex fail with that same error.

vacuum is run regularly via a cron job.

-jason
On Feb 14, 2004, at 2:29 PM, Joshua D. Drake wrote:

Hello,

When was the last time you ran a reindex? Or a vacuum / vacuum full?

Sincerely,

Joshua D. Drake

On Sat, 14 Feb 2004, Jason Essington wrote:

I am running PostgreSQL 7.3.3 on OS X Server 10.2

The database has been running just fine for quite some time now, but
this morning it began pitching the error:
ERROR: cannot read block 176 of tfxtrade_details: Numerical result
out of range
any time the table tfxtrade_details is accessed.

A description of the table is at the end of this email

I have a backup from last night, so I haven't lost much data (if any),
but I am curious if there is a way to recover from this (beyond
restoring from backup) and how I would go about figuring out what
caused it to prevent it from happening again.

I will keep a copy of the data directory if anyone wants me to do any
analysis on it (I will need instructions).

Any insights would be appreciated.

Thanks

Jason Essington
jaessing@greenrivercomputing.com

hedgehog=# \d tfxtrade_details
Table "public.tfxtrade_details"
Column | Type | Modifiers
---------------+--------------------------+-----------
rid | integer | not null
clientid | integer |
tradeid | integer |
rollid | integer |
rollpct | numeric(10,8) |
expdetailid | integer |
expid | integer |
contractpct | numeric(10,8) |
contractamt | numeric(18,2) |
origpct | numeric(10,8) |
origamt | numeric(18,2) |
acctgperiod | integer |
acctgperiodid | integer |
editdate | timestamp with time zone |
edituserid | character varying(48) |
parentid | integer |
entityid | integer |
tradedate | date |
maturitydate | date |
strategyid | integer |
currencyid | integer |
Indexes: tfxtrade_details_pkey primary key btree (rid),
tfxlinks_entityid_index btree (entityid),
tfxlinks_expdetailid_index btree (expdetailid),
tfxlinks_expid_index btree (expid),
tfxlinks_mdate_index btree (maturitydate),
tfxlinks_parentid_index btree (parentid),
tfxlinks_strategy_index btree (strategyid),
tfxlinks_tradeid_index btree (tradeid)
Triggers: RI_ConstraintTrigger_30891,
RI_ConstraintTrigger_30894,
tfxdetail_delete_trigger

---------------------------(end of
broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faqs/FAQ.html

--
Co-Founder
Command Prompt, Inc.
The wheel's spinning but the hamster's dead

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster

--
Co-Founder
Command Prompt, Inc.
The wheel's spinning but the hamster's dead

Starting in single user mode and reindexing the database didn't fix the
error, although it seemed to run just fine.

Vacuum verbose ran until it hit the tfxtrade_details table and then it
died with that same error. it didn't whine about any other problems
prior to dying.

INFO: --Relation public.tfxtrade_details--
ERROR: cannot read block 176 of tfxtrade_details: Numerical result out
of range

Guess there is just something really munged in this one. I'll just try
to restore it from the backup.

Interesting, when I went to copy my data directory out of the way, I
received this from cp:

cp: data/base/16976/17840: Result too large

might be a clue

-jason

On Feb 14, 2004, at 5:01 PM, Joshua D. Drake wrote:

Interesting, when I went to copy my data directory out of the way, I
received this from cp:

cp: data/base/16976/17840: Result too large

might be a clue

I don't think it's PostgreSQL. I would suggest unmounting the volume and
running fsck (or the equivalent for your environment.)

If it's a volumne you can't unmount while the system is running, but you are
running Linux, you could boot a LiveCD distribution of some sort (Gentoo,
Knoppix, and others) and fsck the partition from there.