cloudstack-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Paul Angus <paul.an...@shapeblue.com>
Subject RE: CloudStack Logging
Date Thu, 21 Apr 2016 07:56:30 GMT
Hi Frank,

'thinking out loud' - ideally I'd like to see this failure (and it reasons) surfaced to the
admins outside of the logs.  Off the top of my head - a (helpful) detailed event/alert notification.

More helpful  info in the message (ie not 'host 24') *should* be pretty simple.  I'll have
a look and ask one of our developers to help me with the actual code, once I've tracked it
down.



Kind regards,

Paul Angus

Regards,

Paul Angus

paul.angus@shapeblue.com 
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue

-----Original Message-----
From: Frank Louwers [mailto:frank@openminds.be] 
Sent: 21 April 2016 08:14
To: users@cloudstack.apache.org
Subject: Re: CloudStack Logging

Hi Paul,

I’d love to see improvement in that area! Especially for my operational admins, the biggest
issue we face is this:

- Someone wants to deploy a new VM, and it fails.

- Currently, it’s very hard to figure out exactly why. The best they get is “Capacity
Planner failure”.

- They come to me :)


I think these failures are quite an easy use-case to improve logging on. If a VM can’t be
started because of a “capacity” issue, it would we good to log (in a non-debug, non-info
log):

- Which hosts were considered for CPU/Mem (based on zone, tags etc) (list them by name pref,
not by “host 24” which means nothing to them as that number isn’t anywhere in the CS
UI)

- Which of those considered hosts have the capacity needed (and for those that are excluded,
weather they are excluded for RAM or CPU reasons)

- Exact same for Storage Pools

This would solve over 95% of my “Frank, I can’t find out why this VM won’t deploy”
logging issues.

Regards,

Frank


> On 20 Apr 2016, at 22:05, Paul Angus <paul.angus@shapeblue.com> wrote:
> 
> Hi,
> 
> I don't think that it's a question of importance. INFO vs DEBUG should be telling you
different things. ERROR and WARN are also quite different.
> In general I'm targeting the cloud operational admins, the people who need to know the
health of their cloud and deal with issues as they're constantly reading the log.
> 
> However, I'm not proposing to add or remove any messages, just revisit the categorisation.
If the 'operator' has their logging on debug they'd actually see the same messages. The idea
is to make turning the logging down to INFO feasible.
> 
> You'd turn the logging back up to pick up code issues rather than just operational issues.
> 
> Some VERY simplified definitions might be:
> 
> DEBUG:  inner workings of CloudStack that only a developer can 'understand'
> INFO:  actions that CloudStack is performing or information analysis (host 5 is full
so not going to use it).
> WARN: this isn't good
> ERROR: something just didn't work.
> 
> 
> Grabbing a couple of examples
> 
> these should be debug - I can't 'do' anything with this information.
> 
> INFO  [o.a.c.f.j.i.AsyncJobManagerImpl] 
> (AsyncJobMgr-Heartbeat-1:ctx-4119d5bc) Begin cleanup expired 
> async-jobs INFO  [o.a.c.f.j.i.AsyncJobManagerImpl] 
> (AsyncJobMgr-Heartbeat-1:ctx-4119d5bc) End cleanup expired async-jobs 
> INFO  [c.c.h.v.r.VmwareResource] (DirectAgentCronJob-2:ctx-3209f65c) 
> Scan hung worker VM to recycle INFO  [c.c.h.v.r.VmwareResource] 
> (DirectAgentCronJob-113:ctx-56d9f9f2) Scan hung worker VM to recycle
> 
> Whereas this should be WARN. I wouldn't want to lose this by switching 
> off DEBUG
> 
> DEBUG [c.c.c.ClusterManagerImpl] (Cluster-Heartbeat-1:ctx-91226be7) 
> Detected management node left, id:2, nodeIP:10.2.0.6 DEBUG 
> [c.c.c.ClusterManagerImpl] (Cluster-Heartbeat-1:ctx-0bccd078) Detected 
> management node left, id:2, nodeIP:10.2.0.6
> 
> 
> 
> 
> Kind regards,
> 
> Paul Angus
> 
> Regards,
> 
> Paul Angus
> 
> paul.angus@shapeblue.com
> www.shapeblue.com
> 53 Chandos Place, Covent Garden, London  WC2N 4HSUK @shapeblue
> 
> -----Original Message-----
> From: Chaz PC [mailto:dreeems4ever@hotmail.com]
> Sent: 20 April 2016 11:32
> To: users@cloudstack.apache.org
> Subject: RE: CloudStack Logging
> 
> Hello,
> I have a question what is the criteria that you are going to follow to decide the log
record importance. Who is the user that you will design the logging process for. Is it for
finding issues with the installation or is it for security breaches or auditing.
> These questions are good to put in mind when designing the logging process.
> 
> 
> Sent from my Samsung Galaxy smartphone.<div> </div><div>
> </div><!-- originalMessage --><div>-------- Original message 
> --------</div><div>From: Paul Angus <paul.angus@shapeblue.com> 
> </div><div>Date: 20/04/2016  1:01 PM  (GMT+04:00) </div><div>To:

> dev@cloudstack.apache.org, users@cloudstack.apache.org 
> </div><div>Subject: RE: CloudStack Logging </div><div> </div>
Hi 
> Simon,
> 
> My gut says that it's probably not worth going back before 4.3.  There have been large
changes since then but I think that it's better to get as much data as possible, limiting
to 4.6+ might not give a particularly large install base.
> 
> 
> Kind regards,
> 
> Paul Angus
> 
> Regards,
> 
> Paul Angus
> 
> paul.angus@shapeblue.com
> www.shapeblue.com
> 53 Chandos Place, Covent Garden, London  WC2N 4HSUK @shapeblue
> 
> -----Original Message-----
> From: Simon Weller [mailto:sweller@ena.com]
> Sent: 19 April 2016 21:51
> To: users@cloudstack.apache.org; dev@cloudstack.apache.org
> Subject: Re: CloudStack Logging
> 
> Paul,
> 
> Are you wanting for focus on logs on releases later than a particular version (e.g. 4.6)?
> 
> - Si
> ________________________________________
> From: Paul Angus <paul.angus@shapeblue.com>
> Sent: Tuesday, April 19, 2016 10:55 AM
> To: users@cloudstack.apache.org; dev@cloudstack.apache.org
> Subject: RE: CloudStack Logging
> 
> No problem.
> 
> Good question - Any and all logs.  Often, it's the mundane stuff which obscures the messages
that you really need to see!
> 
> 
> Kind regards,
> 
> Paul Angus
> 
> Regards,
> 
> Paul Angus
> 
> paul.angus@shapeblue.com
> www.shapeblue.com
> 53 Chandos Place, Covent Garden, London  WC2N 4HSUK @shapeblue
> 
> -----Original Message-----
> From: williamstevens@gmail.com [mailto:williamstevens@gmail.com] On 
> Behalf Of Will Stevens
> Sent: 19 April 2016 16:53
> To: dev@cloudstack.apache.org
> Cc: users@cloudstack.apache.org
> Subject: Re: CloudStack Logging
> 
> Thanks Paul.  Are you looking for only logs with errors or are you just looking for just
any management server logs?
> 
> *Will STEVENS*
> Lead Developer
> 
> *CloudOps* *| *Cloud Solutions Experts
> 420 rue Guy *|* Montreal *|* Quebec *|* H3J 1S6 w cloudops.com *|* tw 
> @CloudOps_
> 
> On Tue, Apr 19, 2016 at 11:45 AM, Paul Angus 
> <paul.angus@shapeblue.com>
> wrote:
> 
>> Thanks Will,
>> 
>> As far as I know anything sensitive is removed or obfuscated (vcenter 
>> password displayed as h**********).  IP addresses and hostnames are shown.
>> Users may want to use a tool like SED to replace IP addresses or 
>> internal domain names
>> 
>> http://forums.qrz.com/index.php?threads/remove-ip-addresses-from-log-
>> f
>> iles.204196/
>> 
>> 
>> Kind regards,
>> 
>> Paul Angus
>> 
>> Regards,
>> 
>> Paul Angus
>> 
>> paul.angus@shapeblue.com
>> www.shapeblue.com
>> 53 Chandos Place, Covent Garden, London  WC2N 4HSUK @shapeblue
>> 
>> -----Original Message-----
>> From: williamstevens@gmail.com [mailto:williamstevens@gmail.com] On 
>> Behalf Of Will Stevens
>> Sent: 19 April 2016 16:40
>> To: dev@cloudstack.apache.org
>> Cc: users@cloudstack.apache.org
>> Subject: Re: CloudStack Logging
>> 
>> I like this initiative Paul.  I just wanted to check with you.  Do 
>> you know if there is any confidential information like credentials or 
>> the like that is logged in the management server log?  I thought we 
>> had done a push at one point to make sure no sensitive data was 
>> logged, but I don't remember for sure.  I think knowing the answer to 
>> this will help people feel more comfortable uploading logs.
>> 
>> Thanks for the valuable initiative...
>> 
>> *Will STEVENS*
>> Lead Developer
>> 
>> *CloudOps* *| *Cloud Solutions Experts
>> 420 rue Guy *|* Montreal *|* Quebec *|* H3J 1S6 w cloudops.com *|* tw 
>> @CloudOps_
>> 
>> On Tue, Apr 19, 2016 at 11:34 AM, Paul Angus 
>> <paul.angus@shapeblue.com>
>> wrote:
>> 
>>> Hi All,
>>> 
>>> I'm running an initiative to improve CloudStack logging.  I'm going 
>>> to be analysing logs and re-categorising the DEBUG, INFO, WARN and 
>>> ERROR where appropriate such that we can reduce the default logging 
>>> level to INFO without losing important troubleshooting information, 
>>> while at the same time making the management-server log far more 
>>> readable for operational admins.
>>> 
>>> To achieve this I need logs to work with.
>>> 
>>> Please could anyone interested upload logs to:
>>> https://shapeblue.brickftp.com
>>> 
>>> The logs will be treated in the strictest confidence and public 
>>> download of the logs will not be available.
>>> I'm especially interested in logs which you found particularly 
>>> difficult to read or specific messages which you think are 
>>> misscategorised
>>> 
>>> If you wish, put comments at the top of the logs or send a message 
>>> directly or via the mailing lists pointing out specific issues.
>>> Please compress the logs!
>>> 
>>> Kind regards,
>>> 
>>> Paul Angus
>>> 
>>> 
>>> Regards,
>>> 
>>> Paul Angus
>>> 
>>> paul.angus@shapeblue.com
>>> www.shapeblue.com
>>> 53 Chandos Place, Covent Garden, London  WC2N 4HSUK @shapeblue
>>> 
>> 

Mime
View raw message