| Title: | AdvFS Support/Info/Questions Notefile |
| Notice: | note 187 is Freq Asked Questions;note 7 is support policy |
| Moderator: | DECWET::DADDAMIO |
| Created: | Wed Jun 02 1993 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 1077 |
| Total number of notes: | 4417 |
Does anyone experience following messages?
"vmunix: msfs_bread: msfs_getpage returned error 5"
System configuration;
Alpha Server 8200, 4cpu
KZPSA-BB
HSZ40 V3.0
Bind volume set(raid 5 volume set x2)
RZ29B x 5 (raid 5)
RZ29B x 5 (raid 5)
Digital UNIX V4.0A
/var/adm/messages;
Mar 24 09:48:47 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Mar 26 00:44:03 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Mar 30 22:29:50 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 3 13:01:24 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 8 11:58:55 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 10 11:21:17 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 13 23:05:48 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 14 12:49:22 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 15 10:58:51 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 23 10:14:28 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 25 02:12:27 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 27 22:52:30 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
Apr 09 10:08:30 oske1 vmunix: msfs_bread: msfs_getpage returned error 5
There are no additional error messages. Only this error line
were put in the messages file.
According to the source code about msfs_getpage, error 5 is EIO.
I know we have to analyze customer running system. But this system is
the mail server, so we cannot stop this production system.
This system had trouble about Bit Map Metadata. We extended BMT
preallocate and extension size and also extend Maximum caching size
30% on Mar 23 then we found out these messages. There were no same
errors before Mar 24.
Does anyone experience this error?
Susumu Suwabe
Japan MCS/CSC/SDO
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1070.1 | Check for hardware errors first | UNIFIX::HARRIS | Juggling has its ups and downs | Thu May 29 1997 07:36 | 27 |
The first thing to do is to check the /var/adm/binary.errlog for any
hardware errors being reported.
If you are getting hardware errors reported, then address the hardware
errors first.
----
You can also try running /sbin/advfs/verify against the domain to see
if there are any reported errors from the domain. (verify requires
that all the filesets in the domain be dismounted, so I suspect this
will have to be scheduled.)
----
When you say
"This system had trouble about Bit Map Metadata. We extended BMT
preallocate and extension size and also extend Maximum caching size
30% on Mar 23 then we found out these messages. There were no same
errors before Mar 24."
Does this mean the customer did a vdump/vrestore of the domain so that
a new domain could be created with larger initial allocation of the BMT
via -p and a larger extent size for future groth via -x ???
Bob Harris
| |||||
| 1070.2 | preallocate, extention size | TKTVFS::SUWABE | Susumu Suwabe /OCC/CSC/MCS | Thu May 29 1997 19:47 | 26 |
We have no hardware errors.
> "This system had trouble about Bit Map Metadata. We extended BMT
> preallocate and extension size and also extend Maximum caching size
> 30% on Mar 23 then we found out these messages. There were no same
> errors before Mar 24."
>> Does this mean the customer did a vdump/vrestore of the domain so that
>> a new domain could be created with larger initial allocation of the BMT
>> via -p and a larger extent size for future groth via -x ???
customer has huge files(2,000,000 files). we use this command.
#mkfdmn -p 288000 -x 10240 -l 2560
And also extend AdvfsCacheMaxPercent from default value to 30%.
We try to verify volume set.
Thanks.
S.Suwabe
| |||||