So, it looks like my problems was that I pasted the wrong path
into the Forest Config: .../blah/Forests, which created a new Forests
directory under Forests, with an empty forest in it. (hanging head in
shame)
So now it's failing to mount because it can't memory map the
Timestamps file. Which means I'll have to load the data into the
VM after all, I suppose (30GB).
---
Ron Hitchens {mailto:[email protected]} Ronsoft Technologies
+44 7879 358 212 (voice) http://www.ronsoft.com
+1 707 924 3878 (fax) Bit Twiddling At Its Finest
"No amount of belief establishes any fact." -Unknown
On Jun 19, 2013, at 10:54 PM, Ron Hitchens <[email protected]> wrote:
>
> No. In the forest status it shows all zeros. But, in the database
> status it shows "unknown" for both Fragments and Deleted Fragments.
>
> ---
> Ron Hitchens {mailto:[email protected]} Ronsoft Technologies
> +44 7879 358 212 (voice) http://www.ronsoft.com
> +1 707 924 3878 (fax) Bit Twiddling At Its Finest
> "No amount of belief establishes any fact." -Unknown
>
> On Jun 19, 2013, at 10:03 PM, Danny Sokolsky <[email protected]>
> wrote:
>
>> Are all of the fragments showing up as deleted fragments by any chance?
>> Look at xdmp:forest-counts.
>>
>> -Danny
>>
>> -----Original Message-----
>> From: [email protected]
>> [mailto:[email protected]] On Behalf Of Michael
>> Blakeley
>> Sent: Wednesday, June 19, 2013 2:01 PM
>> To: MarkLogic Developer Discussion
>> Subject: Re: [MarkLogic Dev General] Mounting an alien forest
>>
>> That should work: x86_64 is x86_64, and I have moved forests around like
>> that.
>>
>> My hunch is that the VM bridge to the host filesystem is the culprit. Maybe
>> you could organize enough disk space inside the VM to copy it over?
>>
>> -- Mike
>>
>> On 19 Jun 2013, at 13:58 , Ron Hitchens <[email protected]> wrote:
>>
>>>
>>> I have a forest (filesystem copy) from a client that I need to mount
>>> and examine on my machine. It comes from a 64 bit RedHat Enterprise
>>> system.
>>>
>>> I built a VM (VirtualBox) and loaded CentOS 64bit onto it and
>>> installed the 64bit CentOS build of MarkLogic 6 inside. I've
>>> successfully mounted the Forest on CentOS (after a false start where
>>> the filesystem was read-only and ML didn't like that) but the forest
>>> status shows that the forest is empty.
>>>
>>> The data all seems to to be present on disk (it resides on the host
>>> filesystem, a Mac) and none of the files seem to have been modified.
>>> I shutdown MarkLogic and re-unpacked the forest data to be sure. I
>>> restarted ML and it happily mounted but still shows empty.
>>>
>>> Both are 64bit Linux systems. Is the difference between RedHat and
>>> CentOS enough to make the forest be unusable between them? I though
>>> it was endian-ness and word length that mattered.
>>>
>>> Any ideas?
>>>
>>> ---
>>> Ron Hitchens {mailto:[email protected]} Ronsoft Technologies
>>> +44 7879 358 212 (voice) http://www.ronsoft.com
>>> +1 707 924 3878 (fax) Bit Twiddling At Its Finest
>>> "No amount of belief establishes any fact." -Unknown
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> General mailing list
>>> [email protected]
>>> http://developer.marklogic.com/mailman/listinfo/general
>>>
>>
>> _______________________________________________
>> General mailing list
>> [email protected]
>> http://developer.marklogic.com/mailman/listinfo/general
>> _______________________________________________
>> General mailing list
>> [email protected]
>> http://developer.marklogic.com/mailman/listinfo/general
>
_______________________________________________
General mailing list
[email protected]
http://developer.marklogic.com/mailman/listinfo/general
