Synchronization using https, posted by Alexandre Lindote on Thu Jan 18 16:29:43 2007
|
Hi,
I have an elog server running on a linux desktop machine inside our institutional firewall. It can be accessed from inside the firewall directly using http, but users from outside have to use https, with apache (and its ssl module) controlling the access.
I also set up another server on my laptop, just for backup purposes, and synchronized it inside the firewall (using a direct http connection). No problems there!
Now we want to have several servers set up (in different institutions), and have them synchronized automatically. For this I installed elog on a mac server (in another institution, so outside our firewall), also accessible only through https (again using apache).
I configured all the required mirror settings, but always get an error: "Remote server is not an ELOG server".
I've tried changing the status of the 2 machines (server-client), but get the same error...
I also tried different approaches for the Mirror server parameter, like
https://the.remote.server
https://the.remote.server:443
http://the.remote.server:443
but no luck...
Running elog with the -D flag off issued this message:
| Quote: | HTTP/1.1 401 Authorization Required
Date: Thu, 18 Jan 2007 15:22:10 GMT
Server: Apache/1.3.33 (Darwin) DAV/1.0.3 mod_ssl/2.8.24 OpenSSL/0.9.7l
WWW-Authenticate: Basic realm="zeplin2"
Connection: close
Content-Type: text/html; charset=iso-8859-1
Authorization Required
This server could not verify that you
are authorized to access the document
requested. Either you supplied the wrong
credentials (e.g., bad password), or your
browser doesn't understand how to supply
the credentials required.
Apache/1.3.33 Server at shalott.pp.rl.ac.uk Port 80
|
The servers have user level passwords, and I'm mirroring with Mirror user parameter set to the administrator...
Any ideas?
Thanks in advance
Alex |
Re: Synchronization using https, posted by Stefan Ritt on Thu Jan 18 20:19:22 2007
|
| Alexandre Lindote wrote: | | Now we want to have several servers set up (in different institutions), and have them synchronized automatically. |
Synchronization works such that the "client" elogd (the one on which the synchronization is started) "emulates" a browser and accesses the other elogd server. Since SSL is not built into elogd, it can only use the http protocol, not https. If you want secure synchronization, the only chance you have is to make a SSH tunnel on which you run the http protocol for synchronization. |
|
Re: Synchronization using https, posted by Marcus Hauser on Tue Feb 13 15:11:18 2007
|
| Stefan Ritt wrote: |
| Alexandre Lindote wrote: | | Now we want to have several servers set up (in different institutions), and have them synchronized automatically. |
Synchronization works such that the "client" elogd (the one on which the synchronization is started) "emulates" a browser and accesses the other elogd server. Since SSL is not built into elogd, it can only use the http protocol, not https. If you want secure synchronization, the only chance you have is to make a SSH tunnel on which you run the http protocol for synchronization. |
Would it be possible to use rsync to syncronise two elogd servers on different servers via secure synchronization on the file level?
E.g.
rsync -av -e ssh /var/lib/elog/logbooks/ elog.remote.de:/var/lib/elog/logbooks/ |
|
Re: Synchronization using https, posted by Stefan Ritt on Tue Feb 13 15:36:08 2007
|
| Marcus Hauser wrote: | Would it be possible to use rsync to syncronise two elogd servers on different servers via secure synchronization on the file level?
E.g.
rsync -av -e ssh /var/lib/elog/logbooks/ elog.remote.de:/var/lib/elog/logbooks/ |
Do do this, you have to do the following:
- Stop both elogd servers on each side
- Do the rsync
- Restart the elogd servers
This is necessary because an elogd server does not rescan the database files unless it's restarted. Note that you only can synchronize on a day level. Like if in a single day new entries are made on both sides, you have a conflict. This is not the case if you do the internal elogd synchronization, because this works on an entry level (actually that was the reason why this internal synchronization was invented). So new entries on both sides which were made as the same day will be merged intelligently. |
ELOG Crash by many email address, posted by An Thai on Fri Jan 12 15:28:46 2007 
|
Hello,
when I try to set above 112 email addresses in
"Email All = ..."
the Elog service will crash with the error ntdll.dll or memory addresses x0000000.
This problem does not happen when I reduce the number of email addresses.
Have you had the same problem? |
|
Re: ELOG Crash by many email address, posted by Stefan Ritt on Tue Jan 16 23:01:30 2007
|
| An Thai wrote: | Hello,
when I try to set above 112 email addresses in
"Email All = ..."
the Elog service will crash with the error ntdll.dll or memory addresses x0000000.
This problem does not happen when I reduce the number of email addresses.
Have you had the same problem? |
I fixed that crash in the current version, but there is actually a limit of 100 email addresses. So anyhow you won't get the 12 ones above 100. I will try to increase that limit in the next version of elog. |
|
Re: ELOG Crash by many email address, posted by Grant Jeffcote on Tue Feb 6 18:19:57 2007
|
| Stefan Ritt wrote: |
| An Thai wrote: | Hello,
when I try to set above 112 email addresses in
"Email All = ..."
the Elog service will crash with the error ntdll.dll or memory addresses x0000000.
This problem does not happen when I reduce the number of email addresses.
Have you had the same problem? |
I fixed that crash in the current version, but there is actually a limit of 100 email addresses. So anyhow you won't get the 12 ones above 100. I will try to increase that limit in the next version of elog. |
Hi Stefan,
I'm also having a problem in the latest version with Elog crashing with multiple email addresses.
I am using 'MOptions' for email list selection using the execute shell script to send a predefined list (ie. the email list is in a text file not in the Elog config).
It seems when the list contains exactly 9 addresses it causes Elog to crash with a failure similar to An's, any less or any more and they are sent fine? Also when certain combinations of multiple list selections are made the Elog process will also crash? I am running the Elog process as an application (ie. not as a service, as per An's workaround) as it's the only way I can get the -x switch to work with Windows, this means the Elog daemon has to be manually restarted. Running WinXP SP2.
Any suggestions?
Thx |
Two different attributes with the same list, posted by deletoille on Wed Jan 17 09:20:17 2007
|
In this picture of my project, All the attributes included in the red circle are the same except the names of the attribute “Equipement Incriminé” and “Equipement impacté”.
“Equipement Incriminé” and “Equipement impacté” have the same list.
The displayed list for the both attributes is the multiple conditions result coming from the choice of “groupe incriminé or impacté, sous ensemble, selection, and localisation”.
So if I have to program the list twice, it will be twice bigger.
Do I need to program the same list twice for the two attributes or is there a simple command to say that the two precedent attributes are equal?
Thanks in advance
Xavier Deletoille |
|
Re: Two different attributes with the same list, posted by Stefan Ritt on Wed Jan 17 12:49:25 2007
|
| deletoille wrote: | In this picture of my project, All the attributes included in the red circle are the same except the names of the attribute “Equipement Incriminé” and “Equipement impacté”.
“Equipement Incriminé” and “Equipement impacté” have the same list.
The displayed list for the both attributes is the multiple conditions result coming from the choice of “groupe incriminé or impacté, sous ensemble, selection, and localisation”.
So if I have to program the list twice, it will be twice bigger.
Do I need to program the same list twice for the two attributes or is there a simple command to say that the two precedent attributes are equal?
|
I don't understand your problem. The picture you attached, was this some ELOG output (it looks very different than the usual one...)? Do mean you have two attributes with the same "Options <attribute> = ...,..,.." list? With "multiple conditions" you mean that you do a search in ELOG with a condition one one attribute, but in the filter list both attributes having the same option list are shown. Is this correct? What section of the configuration file are you talking about?
Best regards
Stefan |
|
Re: Two different attributes with the same list, posted by deletoille on Wed Jan 17 13:27:46 2007
|
| Stefan Ritt wrote: |
| deletoille wrote: | In this picture of my project, All the attributes included in the red circle are the same except the names of the attribute “Equipement Incriminé” and “Equipement impacté”.
“Equipement Incriminé” and “Equipement impacté” have the same list.
The displayed list for the both attributes is the multiple conditions result coming from the choice of “groupe incriminé or impacté, sous ensemble, selection, and localisation”.
So if I have to program the list twice, it will be twice bigger.
Do I need to program the same list twice for the two attributes or is there a simple command to say that the two precedent attributes are equal?
|
I don't understand your problem. The picture you attached, was this some ELOG output (it looks very different than the usual one...)? Do mean you have two attributes with the same "Options <attribute> = ...,..,.." list? With "multiple conditions" you mean that you do a search in ELOG with a condition one one attribute, but in the filter list both attributes having the same option list are shown. Is this correct? What section of the configuration file are you talking about?
Best regards
Stefan |
I'm soory, i'm very poor in english.
Here is a very small part of my program:
{AE&Anneau&Panne} Moptions Défaut = Flowmeter or Door meter, Communication, Earth, IAC, Driver, Temp. Ambiant, Phase, Cooling Water, Overvoltage, OverCurrent, Sample Card, Ripple, EEPROM, Autres
{AE&Anneau&Gr} Moptions Equipement incriminé = Q1, Q2, Q3, Q4, Q5, Q6, Q7, Q8, Q9, Q10, QT, S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, Alim correcteurs H., Alim correcteurs V., alim. Dipole, Alim Q., Alim S.
{AE&Anneau&Se&C01} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C02} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.2-CH, S8.2-QT, S9.1-CV, S9.2-CV, S10.1-CH, S10.1-QT, S10.2-CH, COR.
{AE&Anneau&Se&C03} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.1-QT, S8.2-CH, S9.1-CV, S9.2-CV, S10.1-CH, S10.2-CH, S10.2-QT, COR.
{AE&Anneau&Se&C04} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C05} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C06} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.2-CH, S8.2-QT, S9.1-CV, S9.2-CV, S10.1-CH, S10.1-QT, S10.2-CH, COR.
{AE&Anneau&Se&C07} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.1-QT ,S8.2-CH, S9.1-CV, S9.2-CV, S10.1-CH, S10.2-CH, S10.2-QT, COR.
{AE&Anneau&Se&C08} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C09} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C10} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.2-CH ,S8.2-QT, S9.1-CV, S9.2-CV, S10.1-CH, S10.1-QT, S10.2-CH, COR.
{AE&Anneau&Se&C11} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.1-QT, S8.2-CH, S9.1-CV, S9.2-CV, S10.1-CH, S10.2-CH, S10.2-QT, COR.
{AE&Anneau&Se&C12} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C13} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
{AE&Anneau&Se&C14} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.2-CH, S8.2-QT, S9.1-CV, S9.2-CV, S10.1-CH, S10.1-QT, S10.2-CH, COR.
{AE&Anneau&Se&C15} Moptions Equipement incriminé = QA, Q6.1, Q7.1, Q8.1, Q9.1, Q10.1, QB, Q6.2, Q7.2, Q8.2, Q9.2, Q10.2, S7.1-CV, S7.2-CV, S8.1-CH, S8.1-QT, S8.2-CH, S9.1-CV, S9.2-CV, S10.1-CH, S10.2-CH, S10.2-QT, COR.
{AE&Anneau&Se&C16} Moptions Equipement incriminé = QA, Q1, Q2, Q3, Q4.1, Q5.1, QB, Q4.2, Q5.2, Q6, Q7, Q8, S1-CH, S1-QT, S2-CV, S4-CH, S4-CV, S4-QT, S5-CV, S6-CH, COR.
Do I have to program the list for "Equipement Impacté" or is there a simple command to say that "Equipement incrimé" and "equipement inpacté" have the same list?
Thanks in advance
Xavier Deletoille |
|
Re: Two different attributes with the same list, posted by Stefan Ritt on Wed Jan 17 13:33:27 2007
|
| deletoille wrote: | | Do I have to program the list for "Equipement Impacté" or is there a simple command to say that "Equipement incrimé" and "equipement inpacté" have the same list? |
Ok, now I understand. No, there is no simple way, you have to specify both lists explicitly. But it's just a copy and paste in your editor 
By the way, have you discovered Google Translate? Like http://translate.google.com/translate?u=http%3A%2F%2Fmidas.psi.ch%2Felogs%2FForum%2F2115&langpair=en%7Cfr
Not perfect, but readable. I know a Korean and a Japanese guy who do not speak English but talk to each other via email with this automatic translation tool Korean->English->Japanese and back. Amazing. |
|
How many topgroups can be defined?, posted by Exaos Lee on Sat Jan 13 17:35:45 2007
|
I defined 6 top groups in my cfg file as the attachment.
But ... It doesn't work as I wanted.
The 4th and 5th groups need login. But when I clicked the 4th or 5th group, the login isn't needed.
And the 4th and 5th password files are not generated automatically. |
|
Re: How many topgroups can be defined?, posted by Stefan Ritt on Wed Jan 17 12:55:07 2007
|
| Exaos Lee wrote: | I defined 6 top groups in my cfg file as the attachment.
But ... It doesn't work as I wanted.
The 4th and 5th groups need login. But when I clicked the 4th or 5th group, the login isn't needed.
And the 4th and 5th password files are not generated automatically. |
When changing groups and to groups, one can sometimes get fooled by old cookies. Delete all your cookies and try again. You can also install the "Tamper Data" extension for Firefox (https://addons.mozilla.org/firefox/966/) which shows you which cookies are sent to your ELOG server. This should reveal the problem. |
|
Addition of a Print CSS, posted by Ian MacDonald on Mon Jan 15 09:56:09 2007
|
Is there any way to change the the embeded style sheet reference to allow a seperate style sheet for printing ?
Default config:-
<link rel="stylesheet" type="text/css" href="default.css">
Would like to configure to :-
<link rel="stylesheet" type="text/css" href="default.css media="screen">
<link rel="stylesheet" type="text/css" href="print.css" media="print">
Thanks |
|
Re: Addition of a Print CSS, posted by Stefan Ritt on Tue Jan 16 21:13:42 2007
|
| Ian MacDonald wrote: | Is there any way to change the the embeded style sheet reference to allow a seperate style sheet for printing ?
Default config:-
<link rel="stylesheet" type="text/css" href="default.css">
Would like to configure to :-
<link rel="stylesheet" type="text/css" href="default.css media="screen">
<link rel="stylesheet" type="text/css" href="print.css" media="print">
Thanks |
I implemented your suggestion. You can use now
CSS = default.css&screen, print.css&print
The new code is in SVN revision 1782 and will be contained in the next release. |
Group definition under top group, posted by Exaos Lee on Sat Jan 13 17:47:45 2007
|
Hi, I think, if we can put the "Group xxx = xxx" under some "global xxxx" sections, the configure file will be more clean. Don't you think so? In fact, if the "Group XXX" sections are put under "global XXX" section, the "Top Group XXX" can be ignored. I also found that some options (e.g. "Logbook dir =") under "global" don't work under "global XXX".
Maybe the configure file can be defined more flexible, such as this way:
[global]
Port = 8080
Top configuration = yes
[global Group1]
Group config = group1.cfg
[global Group2]
Group config = group2.cfg
The "group1.cfg" looks like this:
[global]
Top configuration = no
Password file = group1.pwd
Group Linux PCs = Red Hat, Debian, Gentoo
Group Windows PCs = 98, ME, XP, Vista
Group Vista = Home, Pro, Media Center
[Red Hat]
...
|
|
Chinese language files updated, posted by Exaos Lee on Wed Jan 10 22:54:56 2007
|
As the attachment.
Sorry for no update long time. |
|
Preset reply date to nothing, posted by herman on Mon Dec 11 18:02:12 2006
|
How do I preset the reply date to an empty box(es)?
In 2.5.x it worked as
Preset on reply Plan Date = ""
gave me an empty date, which is what I wanted. Now in 2.6.3 I get some date in there (...1969) and have to
manually set it to empty again ?
Thanks |
|
Re: Preset reply date to nothing, posted by Stefan Ritt on Tue Dec 12 15:45:48 2006
|
> How do I preset the reply date to an empty box(es)?
>
> In 2.5.x it worked as
>
> Preset on reply Plan Date = ""
>
> gave me an empty date, which is what I wanted. Now in 2.6.3 I get some date in there (...1969) and have to
> manually set it to empty again ?
>
> Thanks
I fixed that in SVN revision 1777 again, so it will be included in the next release. |
|