On 24/06/15 16:02, Tom Robinson wrote:
> I have a DLE that I'm not sure about when backup up windows clients using
> zmanda.
>
> If the DLE has spaces in it, do I need to escape them? Also, do I need to
> escape special charactes?
>
> e.g.
>
> mito cirris "C:/Program Files (x86)/Cirris" {
>
I believe the above is correct but the planner still fails.
The amanda server is quite old (2.6.0p2).
Using the following DLE in the disklist:
mito cirris "C:/Program Files (x86)/Cirris" {
zwc-compress
}
Fails at planner stage but this:
# /usr/libexec/amanda/planner dailytest
---8<---
GETTING ESTIMATES...
got a bad message, stopped at:
----
"C:/Program Files (x86)/Cirris"----
error result for host mito disk cirris: badly formatted response from mito
error result for host mito disk "C:/TNT": badly formatted response from mito
error result for host mito disk "C:/Startrak": badly formatted response from
mito
planner: time 0.420: getting estimates took 0.389 secs
FAILED QUEUE:
0: mito cirris
1: mito "C:/TNT"
2: mito "C:/Startrak"
DONE QUEUE: empty
ANALYZING ESTIMATES...
planner: FAILED mito cirris 20150630083237 0 "[badly formatted response from
mito]"
planner: FAILED mito "C:/TNT" 20150630083237 0 "[badly formatted response from
mito]"
planner: FAILED mito "C:/Startrak" 20150630083237 0 "[badly formatted response
from mito]"
INITIAL SCHEDULE (size 2064):
---8<---
Interestingly, when I remove the 'diskname' from the DLE the error goes away.
So with a DLE of:
mito "C:/Program Files (x86)/Cirris" {
zwc-compress
}
I see this:
# /usr/libexec/amanda/planner dailytest
---8<---
GETTING ESTIMATES...
planner: time 0.339: got result for host mito disk "C:/Program Files
(x86)/Cirris": 0 -> 23438K, -1 -> -2K, -1 -> -2K
planner: time 0.339: got result for host mito disk "C:/TNT": 0 -> 22552K, -1 ->
-2K, -1 -> -2K
planner: time 0.339: got result for host mito disk "C:/Startrak": 0 -> 369740K,
-1 -> -2K, -1 -> -2K
planner: time 0.339: getting estimates took 0.338 secs
FAILED QUEUE: empty
DONE QUEUE:
0: mito "C:/Program Files (x86)/Cirris"
1: mito "C:/TNT"
2: mito "C:/Startrak"
ANALYZING ESTIMATES...
pondering mito:"C:/Program Files (x86)/Cirris"... next_level0 -16616 last_level
-1 (due for level 0) (new disk, can't switch to degraded mode)
curr level 0 nsize 23438 csize 11719 total size 14815 total_lev0 11719
balanced-lev0size 1674
pondering mito:"C:/TNT"... next_level0 -16616 last_level -1 (due for level 0)
(new disk, can't switch to degraded mode)
curr level 0 nsize 22552 csize 11276 total size 27123 total_lev0 22995
balanced-lev0size 3284
pondering mito:"C:/Startrak"... next_level0 -16616 last_level -1 (due for level
0) (new disk, can't switch to degraded mode)
curr level 0 nsize 369740 csize 184870 total size 213025 total_lev0 207865
balanced-lev0size 29694
INITIAL SCHEDULE (size 213025):
mito "C:/Startrak" pri 16617 lev 0 nsize 369740 csize 184870
mito "C:/Program Files (x86)/Cirris" pri 16617 lev 0 nsize 23438 csize 11719
mito "C:/TNT" pri 16617 lev 0 nsize 22552 csize 11276
---8<---
The syntax for a DLE from the documentation is this:
hostname diskname [diskdevice] dumptype [spindle [interface] ]
But using a diskname fails at planner stage as shown in the first example. Also
the the DLE with a diskname disrupts the other DLEs that don't have disknames.
That is, the following also fail until I remove the diskname from the other DLE.
mito "C:/Startrak" {
#starttime 0525
zwc-compress
}
mito "C:/TNT" {
#starttime 0525
zwc-compress
}
Is it because the amanda version I'm running is very old or is there a problem
elsewhere? Maybe I just did something wrong.
Kind regards,
Tom
signature.asc
Description: OpenPGP digital signature
