Re: Weird failures on lorikeet - Mailing list pgsql-hackers

From Thomas Munro
Subject Re: Weird failures on lorikeet
Date
Msg-id CAEepm=3AQg2REZYiJbni7NPsD3acn5fXS+By3E10BGiGwGSa5A@mail.gmail.com
Whole thread Raw
In response to Weird failures on lorikeet  (Andres Freund <andres@anarazel.de>)
Responses Re: Weird failures on lorikeet  (Alvaro Herrera <alvherre@2ndquadrant.com>)
List pgsql-hackers
On Thu, Feb 22, 2018 at 7:06 AM, Andres Freund <andres@anarazel.de> wrote:
> Hi Andrew,
>
> I noticed your animal lorikeet failed in the last two runs:
> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=lorikeet&dt=2018-02-21%2009%3A47%3A17
> TRAP: FailedAssertion("!(((PageHeader) (page))->pd_special >= (__builtin_offsetof (PageHeaderData, pd_linp)))", File:
"/home/andrew/bf64/root/HEAD/pgsql/src/include/storage/bufpage.h",Line: 313)
 
>
> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=lorikeet&dt=2018-02-20%2012%3A46%3A17
> TRAP: FailedAssertion("!(PMSignalState->PMChildFlags[slot] == 1)", File:
"/home/andrew/bf64/root/HEAD/pgsql.build/../pgsql/src/backend/storage/ipc/pmsignal.c",Line: 229)
 
> 2018-02-20 08:07:14.054 EST [5a8c1c3b.21d0:3] LOG:  select() failed in postmaster: Bad address
> 2018-02-20 08:07:14.073 EST [5a8c1c3b.21d0:4] LOG:  database system is shut down
>
> The difference between the last successfull and the last failing build
> is a single comment typo commit.
>
> It kinda looks like there might be some underlying issue on the machine
> with shared memory going away or such?

Bad memory?  Assorted output from recent runs:

+ ERROR:  index "pg_toast_28546_index" contains corrupted page at block 1

TRAP: FailedAssertion("!(PMSignalState->PMChildFlags[slot] == 1)",
File: "/home/andrew/bf64/root/HEAD/pgsql.build/../pgsql/src/backend/storage/ipc/pmsignal.c",
Line: 229)

TRAP: FailedAssertion("!(entry->trans == ((void *)0))", File:
"/home/andrew/bf64/root/HEAD/pgsql.build/../pgsql/src/backend/postmaster/pgstat.c",
Line: 871)

-- 
Thomas Munro
http://www.enterprisedb.com


pgsql-hackers by date:

Previous
From: Thomas Munro
Date:
Subject: Re: 2018-03 CFM
Next
From: Andres Freund
Date:
Subject: Re: [HACKERS] Restrict concurrent update/delete with UPDATE ofpartition key