Re: Parallel Discovery?

2014-03-16 Thread Gaofeng
We tried parallel discovery and login, and found a problem.
 
When discovering, the open-iscsi will generate all the info files in 
/etc/iscsi/nodes. 
If there are same files, it will delete olds and make news.  

Consider the following case: 
  when you discovery secondly, after you delete old info files(but 
do not create news), then you login, and it will failed.

How we do: 
discovery process:

 lock 
 iscisi discovery 
if (exist(/etc/iscsi/nodes/XX)  then 
unlock

login process:

while (failed = login()) { sleep 1}



在 2013年10月24日星期四UTC+8上午4时14分30秒,The Lee-Man写道:

 Hi Mike:

 I have a configuration where discovery is taking a long time because one 
 of the interfaces is disconnected.

 Since it is not possible to reliably detect when an interface is 
 connected, I thought the best way to speed up discovery in such cases would 
 be for discovery to take place in parallel.

 In looking at the code, it seems like node discovery could be done in 
 parallel with a little work. In searching the archives, I found a thread 
 where it looks like you wanted to make discovery in parallel, but that was 
 a while ago.

 Before I start hacking on the code to see what I can do, I wondered if you 
 have any ideas on the best way to do this, or even if it is a good idea at 
 all.

 With only 2 or 3 interfaces, this is not a big deal, but I could see a 
 case where a dozen interfaces could mean a very slow discovery process.
 --
 Lee Duncan



-- 
You received this message because you are subscribed to the Google Groups 
open-iscsi group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to open-iscsi+unsubscr...@googlegroups.com.
To post to this group, send email to open-iscsi@googlegroups.com.
Visit this group at http://groups.google.com/group/open-iscsi.
For more options, visit https://groups.google.com/d/optout.


Parallel Discovery?

2013-10-23 Thread The Lee-Man
Hi Mike:

I have a configuration where discovery is taking a long time because one of 
the interfaces is disconnected.

Since it is not possible to reliably detect when an interface is connected, 
I thought the best way to speed up discovery in such cases would be for 
discovery to take place in parallel.

In looking at the code, it seems like node discovery could be done in 
parallel with a little work. In searching the archives, I found a thread 
where it looks like you wanted to make discovery in parallel, but that was 
a while ago.

Before I start hacking on the code to see what I can do, I wondered if you 
have any ideas on the best way to do this, or even if it is a good idea at 
all.

With only 2 or 3 interfaces, this is not a big deal, but I could see a case 
where a dozen interfaces could mean a very slow discovery process.
--
Lee Duncan

-- 
You received this message because you are subscribed to the Google Groups 
open-iscsi group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to open-iscsi+unsubscr...@googlegroups.com.
To post to this group, send email to open-iscsi@googlegroups.com.
Visit this group at http://groups.google.com/group/open-iscsi.
For more options, visit https://groups.google.com/groups/opt_out.


Re: Parallel Discovery?

2013-10-23 Thread Mike Christie
On 10/23/2013 03:14 PM, The Lee-Man wrote:
 Hi Mike:
 
 I have a configuration where discovery is taking a long time because one
 of the interfaces is disconnected.
 
 Since it is not possible to reliably detect when an interface is
 connected, I thought the best way to speed up discovery in such cases
 would be for discovery to take place in parallel.
 
 In looking at the code, it seems like node discovery could be done in
 parallel with a little work. In searching the archives, I found a thread
 where it looks like you wanted to make discovery in parallel, but that
 was a while ago.
 
 Before I start hacking on the code to see what I can do, I wondered if
 you have any ideas on the best way to do this, or even if it is a good
 idea at all.
 
 With only 2 or 3 interfaces, this is not a big deal, but I could see a
 case where a dozen interfaces could mean a very slow discovery process.

It sounds like a good idea.

-- 
You received this message because you are subscribed to the Google Groups 
open-iscsi group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to open-iscsi+unsubscr...@googlegroups.com.
To post to this group, send email to open-iscsi@googlegroups.com.
Visit this group at http://groups.google.com/group/open-iscsi.
For more options, visit https://groups.google.com/groups/opt_out.