Did it again. It seemed that Google bot came to the elog and tried to obtain "http://nodus.ligo.caltech.edu:8080/robots.txt". That was the last of the log.
Bot came from the AJW's homepage. Also Google FeedFecther came to the elog.

I deleted the yellow box which showed up by default when making an elog entry. Would be nice if we could make it so that you have to click a button to 'opt-in' for the yellow box rather than get it by default.

I added a 'robots.txt' file to the /users/public_html/ area using Google's instructions (it only works with robot compliant crawlers), but am not sure how to put robots.txt into the elog port.

Suresh killed the elogd again from India. This was the log file:

Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7
Keep-Alive: 115
Referer: http://www.ligo.caltech.edu/~ajw/40m_upgrade.html
Cookie: elmode=threaded; __utma=65601905.411937803.1291369887.1291369887.1291369887.1; __utmz=65601905.1291369887.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none); SITESERVER=ID=4981c5fd42ae53c9c9e0980f2072be4f
Via: 1.1 sfire5.du.ac.in:3128 (squid/2.6.STABLE6)
X-Forwarded-For: 10.11.1.120
Cache-Control: max-age=259200
Connection: keep-alive

Stop killing our elog!

After another elog crash, I've blacklisted the domain that Suresh is using by editing the apache httpd.conf. Let's see what happens now.

ELOG has crashed and I restarted it.

Actually the filtering is not effective so far as elog is not using apache but has its own web server inside.
So this just block the access to port 30889 (=SVN, Dokuwiki, etc).

Sorry folks!  I couldnt get to the elog and didnt know that the elog was crashing every time I tried to access it. 

But have found other means to access it and the elog is safe for now!

This was the error today:

GET /40m/ HTTP/1.1
Host: nodus.ligo.caltech.edu:8080
User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.3) Gecko/20100401 Firefox/3.6.3
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8
Accept-Language: en-us,en;q=0.5
Accept-Encoding: gzip,deflate
Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7
Keep-Alive: 115
Connection: keep-alive
Referer: http://www.ligo.caltech.edu/~ajw/40m_upgrade.html
Cookie: elmode=threaded; __utma=65601905.411937803.1291369887.1291369887.1291369887.1; __utmz=65601905.1291369887.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none); SITESERVER=ID=4981c5fd42ae53c9c9e0980f2072be4f

Found exactly the same error messages at the end of the log file.

Google bot  crashed the elog again. Then, I found that Google bot (and I) can crash elogd by trying to show the threaded view.
There looks similar issue reported to the elog forum, the author did not think this is a true bag.

Note: This happens only for the 40m elog. The other elogs (ATF/PSL/TCS/SUS/Cryo) are OK for the threaded view.

 Restarted the elog with the script

Fixed the 40m elog crashing with the threaded display.

This morning I found that Google bot crashed the elog again. I started the investigation and found the threaded mode is fine if we use the recent 10 entries.

I gradually copied the old entries to a temporary elog and found that a deleted elog entry on August 6 had a corrupted remnant in the elog file. This kept crashed the threaded mode.

Once this entry has been eliminated again, the threaded mode got functional.

I hope this eliminates those frequent elog crashing.

Elog wasn't responding at around 10 am this morning.  I killed the elogd process, then used the restart script.

 I restarted the elog using the script.

 again

 The elog was dead this morning. I reanimated it. It is now undead.

with script

Restarted elog.

Restarted the elog with the script.

 with script

Aka, from a hotel in Pisa.

 Restarted Thu May 19 00:21:49 2011 to recover from Jenne's Italian terrorism.

Elog crashed a couple times, restarted it a couple times.

like the subject says...

I restarted the elog with the script as it was not up when I tried to make a post. It was again unresponsive when I went to submit, but this time the script couldn't restart it. The log said it couldn't bind to 8080, which usually happens if the daemon is still running. I pkilled it, then reran the script, and it appears to be working.

Elog crashed / dormant for long time.  A look at the log file indicated that it was busy generating png thumbnails for pdf files.

Restarted at Wed Sep  7 20:41:13 PDT 2011

WTF?

Elog went nonreponsive. SSH'ed into nodus to run restart script. Elog came back ~15 minutes later.

Elog did not respond despite running the /cvs/cds/caltech/elog/start-elog.csh  script two times.  

It worked the after the third restart. 
 

Elog was hanging, restarted with the script.

Elog restart script killed the elog, but didn't restart it.  Restarted by hand.

I've noticed that it always takes running the script twice for it to actually work. I think there's something wrong with how it's doing it. I'll mess with it sometime the elog isn't getting much use.

Elog was not responding and was restarted.

Elog suspended 2 times for 1 hour. Too high frequency.

Restarted.

Elog was hanging. Restarted it with script.

Basically, elog hangs up by the visit of googlebot.

Googlebot repeatedly tries to obtain all (!) of entries by specifying  "page0" and "mode=full".
Elogd seems not to have the way to block an access from the specified IP.
We might be able to use http-proxy via apache. (c.f. http://midas.psi.ch/elog/adminguide.html#secure)

Or can we tweak the source of elod such that it does not accept "page0"?


GET /40m/page0?mode=full&new_entries=1 HTTP/1.1
Host: 131.215.115.52:8080
Connection: Keep-alive
Accept: */*
From: googlebot(at)googlebot.com
User-Agent: Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)
Accept-Encoding: gzip,deflate

 There are much more simple ways to prevent page indexing by googlebot: http://www.google.com/support/webmasters/bin/answer.py?answer=156449

However, I really think that's a less-than-idea solution to get around the actual problem which is that the elog software is a total piece of crap.  If google does not index the log then it won't appear in google search results.

But if there is one url that when requested causes the elog to crash then maybe it's a better solution to cut of just that url.

- elogd itself is a sort of web server which has no freedom to put our own file, we can not put robots.txt

- If we include elog using proxy in the usual web tree of Apache, we can put robots.txt at the root.

- So far, if we prevent "page0" browse by google, we will be saved for a while.

- It seems that the indexing is set to be refused by the following meta tags. But it does not prohibit googlebot to use "page0" URL, of course.

<META NAME="ROBOTS" CONTENT="NOINDEX, NOFOLLOW">

 with script.

Gave the elog its afternoon / tea-time reboot, since it was hanging yet again.

Tried the script 3 times and it didn't come back. Pkill'd and then scripted. That worked.

I modified elogd.c as shown below in order not to allow to display all of the entries at once by specifying page number "0".

After this modification, elog seems to have survived 66 times of attacks by googlebots.

==============================

rossa:src>pwd
/cvs/cds/caltech/elog/elog-2.8.0/src

rossa:src>diff elogd.c_orig elogd.c
19912a19913,19917
>
>    /* KA */
>    if (page_n<1)
>      page_n = 1;
>


On the plus side, it was the first time I've had to do it in a while..

The elog was not responding. I attempted to restart it by running .../start-elog.csh, but this didn't work (even after the usual ~2 times it needs).

Somehow pkill did not kill the daemon, so I kill -9'd it and that did the trick. I ran the start script once more and it came back online.

 Aim: to scan the cavitymodes of IMC

The circuit used: 

 Attachment4

Results obtained:

Attachment 1,2,3

I deleted a bunch of useless categories from the 40m elogd.cfg file. IF you find your category has been deleted, you can now learn how to categorize yourself into the existing categories. ELOG restarted and log file zeroed.