Tue Oct 5 18:35:08 PDT 2004
- Previous message: [Slony1-general] slony causes postgresql children to die
- Next message: [Slony1-general] slony causes postgresql children to die
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Brad Hilton wrote: | hannu at skype.net wrote: | |>> Hello, |>> |>> I've spent the last few days learning slony and setting it up on some |>> test databases with good results. However, when I set up replication on |>> my live database, the slon processes startup and appear to be preparing |>> to copy things over when I get: |>> |>> "The postmaster has commanded this server process to roll back the |>> current transaction and exit, because another server process exited |>> abnormally and possibly corrupted shared memory." |>> |>> This happens continually until I uninstall the slony schema from my |>> database. |> |> Does slon need to be running or is it happening even without it |> (maybe as |> a result of slony triggers doing something bad) ? | | No, slon does not need to be running for the crashes to occur. That's useful to know; that suggests that the reason for the problem may not directly relate to Slony-I. You should consider the possibility that there is some independent problem that is the _real_ thing that needs to be found. It would be unfortunate to direct your attentions all at Slony-I, only to discover that the real problem is something else. We ran into that recently in the context of performance; there was a system behaving strangely, and at "first blush," all fingers were pointing at Slony-I as the thing that had been most recently introduced. ~ Ultimately, the investigations found that there was a bug/misconfiguration in a hardware monitoring process that was related only insomuch as it caused wacky I/O patterns that made COMMITs run slowly, which particularly injured certain Slony-I transactions :-(. The point is that you need to have a pretty high degree of skepticism as to what is the actual cause. Symptoms != Causes. Unfortunately, the more components you glue together, the more things there are that can cause "bad things." :-( | I'll work on getting a stack trace. That should be helpful. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.5 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFBYttECVn6LJfHIAIRAlX0AJ4mYtr/SvEpMbXZNaPFOr7kg2KIywCeMAbN iQbtPe3EtZBEUZn1ako4aMc= =DQ+j -----END PGP SIGNATURE-----
- Previous message: [Slony1-general] slony causes postgresql children to die
- Next message: [Slony1-general] slony causes postgresql children to die
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list