[one-users] Breakage when upgrading from 4.38 to 4.4 ... [Resolved]

[Gareth Bult]

Ok, not sure what the difference is, but this is a fix; 

kvm-probes.d/monitor_ds.sh 


--- monitor_ds.sh.orig 2013-12-05 17:06:35.251763742 +0000 
+++ monitor_ds.sh 2013-12-05 16:55:11.699743037 +0000 
@@ -31,6 +31,10 @@ 
TOTAL_MB=$(df -B1M -P $dir 2>/dev/null | tail -n 1 | awk '{print $2}') 
FREE_MB=$(df -B1M -P $dir 2>/dev/null | tail -n 1 | awk '{print $4}') 

+ USED_MB=${USED_MB:-"0"} 
+ TOTAL_MB=${TOTAL_MB:-"0"} 
+ FREE_MB=${FREE_MB:-"0"} 
+ 
if [ -n "$LVM_SIZE_CMD" ]; then 
LVM_SIZE=$($LVM_SIZE_CMD ${LVM_VG_PREFIX}${ds} 2>/dev/null) 
LVM_STATUS=$? 
The host monitor seems to be calling a generic monitor which is also grabbing information about datastores. 
In an environment (such as VDC) where the data exists remotely, the standard monitor (which uses 'du') isn't 
going to provide any useful results, indeed in this instance it errors and the script returns blanks, which breaks 
the host monitoring. Returning '0' on failure fixes the issue ... 

Any chance this could inserted in the next release please? 


-- 
	Gareth Bult 
“The odds of hitting your target go up dramatically when you aim at it.” 



----- Original Message -----

From: "Gareth Bult" <gareth at linux.co.uk> 
To: users at lists.opennebula.org 
Sent: Thursday, 5 December, 2013 3:47:12 PM 
Subject: [one-users] Breakage when upgrading from 4.38 to 4.4 ... 

Hi, I've run the upgrade just to see if a number of issues I was having were fixed, and I seem to have 
ended up with a fairly terminal issue I can't spot. 

Host monitoring is now failing for hosts that previously seemed to be working. The reason seems to be 
that the host is somehow looking at the datastores, and this in turn is failing. 

The datastore monitoring is correct. 
The error in oned.log is as follows; 


Thu Dec 5 15:24:47 2013 [ONE][E]: Error parsing host information: syntax error, unexpected EQUAL, expecting COMMA or CBRACKET at line 30, columns 494:497. Monitoring information: 
ARCH=x86_64 
MODELNAME="AMD Phenom(tm) II X6 1100T Processor" 
HYPERVISOR=kvm 
TOTALCPU=600 
CPUSPEED=3300 
TOTALMEMORY=16175216 
USEDMEMORY=399320 
FREEMEMORY=15775896 
FREECPU=600.0 
USEDCPU=0.0 
NETRX=33791417902 
NETTX=25290741468 
DS_LOCATION_USED_MB=2637 
DS_LOCATION_TOTAL_MB=40190 
DS_LOCATION_FREE_MB=28202 
DS = [ 
ID = 0, 
USED_MB = 29, 
TOTAL_MB = 40190, 
FREE_MB = 28202 
] 
DS = [ 
ID = 1, 
USED_MB = 2609, 
TOTAL_MB = 40190, 
FREE_MB = 28202 
] 
DS = [ 
ID = 107, 
USED_MB = , 
TOTAL_MB = , 
FREE_MB = 
] 
DS = [ 
ID = 109, 
USED_MB = , 
TOTAL_MB = , 
FREE_MB = 
] 
DS = [ 
ID = 114, 
USED_MB = 1, 
TOTAL_MB = 40190, 
FREE_MB = 28202 
] 
DS = [ 
ID = 2, 
USED_MB = 1, 
TOTAL_MB = 40190, 
FREE_MB = 28202 
] 
HOSTNAME=node1 
VM_POLL=YES 
VERSION="4.4.0" 
So it would appear that it's calling datastore/monitor, and it's failing .. however the script does work and is reporting 
the correct size in the datastore monitor. I've added syslog to the monitor script, it is being called and it is returning 
the correct result ... 

If I call the script on the command line with a pasted in base64 and host it, this too gives the correct results; 

oneadmin at nebula:~/remotes$ ssh node1 /var/lib/one/remotes/datastore/vdc/monitor "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" 109 


TOTAL_MB=1500101.21 
FREE_MB=963230.30 
USED_MB=536871 
Short of trying to dig through oned's source code I'm stuck - can anyone help? 

My datastore driver type is "vdc", even if I insert a static script (I've also tried the monitor driver from "fs") 
that just prints out "TOTAL_MB=0" etc .. I still get the same issue ... 

??? 
(and it was working fine .. before the upgrade??) 


-- 
	Gareth Bult 
“The odds of hitting your target go up dramatically when you aim at it.” 




_______________________________________________ 
Users mailing list 
Users at lists.opennebula.org 
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20131205/070cd498/attachment-0002.htm>