Dear all,

I would like to ask some issues regarding Robinhood.

Our installation consists of ::

SLES.11.1 x86_64
robinhood-tmp_fs_mgr-2.3.1-1.lustre1.8
MySQL-server-community-5.1.49-1.sles11
lustre-1.8.4-2.6.32.59_0.3_default

We are using robinhood for monitoring a number of small (home) FSs and purging 
old files from a large scratch FS.

Lately we have noticed that robinhood takes quite a long time to parse several 
filesystems. For example on the following FS it required ~18 hours in the last 
2 scans.

# lfs df -i | grep summary | grep FS
filesystem summary:    222046620     7534472   214512148   3% /lustre/FS
# lfs df -h | grep summary | grep FS
filesystem summary:        28.7T       13.9T       13.3T  48% /lustre/FS


Personally I think that it occurs mainly because of users’ misbehaviour 
(creating folders with more than 10K files) but I would like your help to 
identify whether there are any other reasons (related to robinhood or the 
database) and ask some questions on how could I fix this.

My idea is ::

To backup all the files/folders related to the robinhood-db.
Delete all the files/folders related to the robinhood-db.
Create a new scan for the problematic FSs (measure the time).
Once this is complete, inform the users from top-dirs to clean up their space 
(archive/delete).
Once this is complete, rescan the FSs (measure the time, check whether 
situation was improved).
Following, act depending on whether the time to scan has improved or not.

Robinhood has been set and monitoring these FSs some years now (the exact time 
is unknown since it was set by a former colleague) and I don’t know if 
accessing a big db (because of old entries) slows down robinhood. As an example 

# du -sh robinhood_SCRATCH/*
4.0K    robinhood_SCRATCH/ACCT_ENTRY_DELETE.TRN
4.0K    robinhood_SCRATCH/ACCT_ENTRY_INSERT.TRN
4.0K    robinhood_SCRATCH/ACCT_ENTRY_UPDATE.TRN
264K    robinhood_SCRATCH/ACCT_STAT.MYD
168K    robinhood_SCRATCH/ACCT_STAT.MYI
12K     robinhood_SCRATCH/ACCT_STAT.frm
56M     robinhood_SCRATCH/ANNEX_INFO.MYD
21M     robinhood_SCRATCH/ANNEX_INFO.MYI
12K     robinhood_SCRATCH/ANNEX_INFO.frm
8.1G    robinhood_SCRATCH/ENTRIES.MYD
456M    robinhood_SCRATCH/ENTRIES.MYI
4.0K    robinhood_SCRATCH/ENTRIES.TRG
12K     robinhood_SCRATCH/ENTRIES.frm
1.6G    robinhood_SCRATCH/STRIPE_INFO.MYD
461M    robinhood_SCRATCH/STRIPE_INFO.MYI
12K     robinhood_SCRATCH/STRIPE_INFO.frm
4.9G    robinhood_SCRATCH/STRIPE_ITEMS.MYD
968M    robinhood_SCRATCH/STRIPE_ITEMS.MYI
12K     robinhood_SCRATCH/STRIPE_ITEMS.frm
4.0K    robinhood_SCRATCH/VARS.MYD
4.0K    robinhood_SCRATCH/VARS.MYI
12K     robinhood_SCRATCH/VARS.frm
4.0K    robinhood_SCRATCH/db.opt

I would like if possible your ideas/comments on my questions above and any 
possible (robinhood-related) advice on how could I further debug why robinhood 
scan takes so long.

Thank you in advance for your time.

Regards,
Stelios

Attachment: smime.p7s
Description: S/MIME cryptographic signature

------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
robinhood-support mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/robinhood-support

Reply via email to